examples/01_skpro_intro.ipynb

{
 "cells": [
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## skpro introduction notebook"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Set-up instructions:** On binder, this should run out-of-the-box.\n",
    "\n",
    "To run locally instead, ensure that `skpro` with basic dependency requirements is installed in your python environment."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`skpro` provides `scikit-learn`-like, `scikit-base` compatible interfaces to:\n",
    "\n",
    "* tabular **supervised regressors with probabilistic prediction modes** - interval, quantile and distribution predictions\n",
    "* **performance metrics to evaluate probabilistic predictions**, e.g., pinball loss, empirical coverage, CRPS\n",
    "* **reductions** to turn non-probabilistic, `scikit-learn` regressors into probabilistic `skpro` regressors, such as bootstrap or conformal\n",
    "* tools for building **pipelines and composite machine learning models**, including tuning via probabilistic performance metrics\n",
    "* symbolic an lazy **probability distributions** with a value domain of `pandas.DataFrame`-s and a `pandas`-like interface\n",
    "\n",
    "**Section 1** provides an overview of common **probabilistic supervised regression workflows** supported by `skpro`.\n",
    "\n",
    "**Section 2** gives an more detailed introduction to **prediction modes, performance metrics, and benchmarking tools**.\n",
    "\n",
    "**Section 3** discusses **advanced composition patterns**, including various ways to add probabilistic capability to any `sklearn` regressor, pipeline building, tuning, ensembling.\n",
    "\n",
    "**Section 4** gives an introduction to how to write **custom estimators** compliant with the `skpro` interface."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "# hide warnings\n",
    "import warnings\n",
    "\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1. Basic probabilistic supervised regression workflows <a class=\"anchor\" id=\"chapter1\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### What is probabilistic supervised regression?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "= supervised regression with probabilistic prediction\n",
    "\n",
    "* given training data `X` (N, d) feature array; N float labels `y`\n",
    "* learn a model that predicts *variation range* of label `y_new`, given `X_new` (N, 1)\n",
    "* precisely: given `X` row, produce conditional distribution for `y`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`skpro` provides estimators for probabilistic supervised regression:\n",
    "\n",
    "* `fit(X, y)` with tabular features `X`, labels `y`, same rows, both `pd.DataFrame`\n",
    "* `predict_interval(X_new, coverage=0.90)` for interval predictions of labels\n",
    "* `predict_quantiles(X_new, alpha=[0.05, 0.95])` for quantile predictions of labels\n",
    "* `predict_var(X_new)` for variance predictions of labels\n",
    "* `predict(X_new)` for mean predictions\n",
    "* `predict_proba(X_new)` for distributional prediction"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.1 basic deployment workflow"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`skpro` regressors are used via `fit` then `predict_proba` etc.\n",
    "\n",
    "Same as `sklearn` regressors - `X` and `y` should be `pd.DataFrame` (`numpy` is also ok but not recommended)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_diabetes\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from skpro.regression.residual import ResidualDouble\n",
    "\n",
    "# step 1: data specification\n",
    "X, y = load_diabetes(return_X_y=True, as_frame=True)\n",
    "X_train, X_new, y_train, _ = train_test_split(X, y)\n",
    "\n",
    "# step 2: specifying the regressor\n",
    "# example - random forest for mean prediction\n",
    "# linear regression for variance prediction\n",
    "reg_mean = RandomForestRegressor()\n",
    "reg_resid = LinearRegression()\n",
    "reg_proba = ResidualDouble(reg_mean, reg_resid)\n",
    "\n",
    "# step 3: fitting the model to training data\n",
    "reg_proba.fit(X_train, y_train)\n",
    "\n",
    "# step 4: predicting labels on new data\n",
    "\n",
    "# probabilistic prediction modes - pick any or multiple\n",
    "# we show the return types in detail below\n",
    "y_pred_proba = reg_proba.predict_proba(X_new)\n",
    "y_pred_interval = reg_proba.predict_interval(X_new, coverage=0.9)\n",
    "y_pred_quantiles = reg_proba.predict_quantiles(X_new, alpha=[0.05, 0.5, 0.95])\n",
    "y_pred_var = reg_proba.predict_var(X_new)\n",
    "y_pred_mean = reg_proba.predict(X_new)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "returns describe predictive conditional distribution of labels\n",
    "\n",
    "details about different return types below"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.1.1 distribution predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`y_pred_proba` is an `skpro` distribution - it has index and columns like `pd.DataFrame`\n",
    "\n",
    "\"we predict that true labels are distributed according to `y_pred_proba`\"\n",
    "\n",
    "(here: distribution marginal by row/columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(<Figure size 640x480 with 5 Axes>,\n",
       " array([<Axes: title={'center': 'target'}, ylabel='3'>,\n",
       "        <Axes: ylabel='197'>, <Axes: ylabel='395'>, <Axes: ylabel='78'>,\n",
       "        <Axes: ylabel='149'>], dtype=object))"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "y_pred_proba = reg_proba.predict_proba(X_new)\n",
    "y_pred_proba.head().plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`skpro` distribution objects are pandas-like"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(111, 1)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_proba.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index([  3, 197, 395,  78, 149, 105, 323,   9, 363, 369,\n",
       "       ...\n",
       "       346, 153, 178, 283,  34, 393, 316, 383, 122,  82],\n",
       "      dtype='int64', length=111)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_proba.index  # same index as X_new"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Index(['target'], dtype='object')"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_proba.columns  # same columns as X_new"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "distribution objects have `sample` and methods such as `mean`, `var`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>181.254107</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>199.392975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>147.874808</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>165.939361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>154.024654</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         target\n",
       "3    181.254107\n",
       "197  199.392975\n",
       "395  147.874808\n",
       "78   165.939361\n",
       "149  154.024654"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_proba.sample().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>163.56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>187.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>126.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>176.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>143.94</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     target\n",
       "3    163.56\n",
       "197  187.08\n",
       "395  126.75\n",
       "78   176.01\n",
       "149  143.94"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_proba.mean().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>524.491525</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>436.051498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>299.445215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>374.212877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>310.055919</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         target\n",
       "3    524.491525\n",
       "197  436.051498\n",
       "395  299.445215\n",
       "78   374.212877\n",
       "149  310.055919"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_proba.var().head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "more details on `skpro` distributions in the \"distributions\" tutorial!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.1.2 interval predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "interval prediction `y_pred_interval` is a `pd.DataFrame`:\n",
    "\n",
    "* rows are the same as `X_new`\n",
    "* columns indicate variables, nominal coverage, and bottom/upper bound\n",
    "\n",
    "\"we predict that value in row falls between bottom/upper with 90% chance\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">target</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">0.9</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>lower</th>\n",
       "      <th>upper</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>125.889924</td>\n",
       "      <td>201.230076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>152.732420</td>\n",
       "      <td>221.427580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>98.286654</td>\n",
       "      <td>155.213346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>144.190993</td>\n",
       "      <td>207.829007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>114.976752</td>\n",
       "      <td>172.903248</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         target            \n",
       "            0.9            \n",
       "          lower       upper\n",
       "3    125.889924  201.230076\n",
       "197  152.732420  221.427580\n",
       "395   98.286654  155.213346\n",
       "78   144.190993  207.829007\n",
       "149  114.976752  172.903248"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_interval = reg_proba.predict_interval(X_new, coverage=0.9)\n",
    "y_pred_interval.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "multiple coverages can be passed:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"6\" halign=\"left\">target</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">0.50</th>\n",
       "      <th colspan=\"2\" halign=\"left\">0.90</th>\n",
       "      <th colspan=\"2\" halign=\"left\">0.95</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>lower</th>\n",
       "      <th>upper</th>\n",
       "      <th>lower</th>\n",
       "      <th>upper</th>\n",
       "      <th>lower</th>\n",
       "      <th>upper</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>148.112984</td>\n",
       "      <td>179.007016</td>\n",
       "      <td>125.889924</td>\n",
       "      <td>201.230076</td>\n",
       "      <td>118.673336</td>\n",
       "      <td>208.446664</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>172.995409</td>\n",
       "      <td>201.164591</td>\n",
       "      <td>152.732420</td>\n",
       "      <td>221.427580</td>\n",
       "      <td>146.152334</td>\n",
       "      <td>228.007666</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>115.078302</td>\n",
       "      <td>138.421698</td>\n",
       "      <td>98.286654</td>\n",
       "      <td>155.213346</td>\n",
       "      <td>92.833832</td>\n",
       "      <td>160.666168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>162.962277</td>\n",
       "      <td>189.057723</td>\n",
       "      <td>144.190993</td>\n",
       "      <td>207.829007</td>\n",
       "      <td>138.095315</td>\n",
       "      <td>213.924685</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>132.063311</td>\n",
       "      <td>155.816689</td>\n",
       "      <td>114.976752</td>\n",
       "      <td>172.903248</td>\n",
       "      <td>109.428161</td>\n",
       "      <td>178.451839</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>393</th>\n",
       "      <td>94.468654</td>\n",
       "      <td>120.131346</td>\n",
       "      <td>76.008663</td>\n",
       "      <td>138.591337</td>\n",
       "      <td>70.014072</td>\n",
       "      <td>144.585928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>316</th>\n",
       "      <td>216.443691</td>\n",
       "      <td>238.596309</td>\n",
       "      <td>200.508608</td>\n",
       "      <td>254.531392</td>\n",
       "      <td>195.333942</td>\n",
       "      <td>259.706058</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>383</th>\n",
       "      <td>94.264854</td>\n",
       "      <td>116.395146</td>\n",
       "      <td>78.345830</td>\n",
       "      <td>132.314170</td>\n",
       "      <td>73.176378</td>\n",
       "      <td>137.483622</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>261.935566</td>\n",
       "      <td>288.344434</td>\n",
       "      <td>242.938827</td>\n",
       "      <td>307.341173</td>\n",
       "      <td>236.769935</td>\n",
       "      <td>313.510065</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>75.425129</td>\n",
       "      <td>88.414871</td>\n",
       "      <td>66.081193</td>\n",
       "      <td>97.758807</td>\n",
       "      <td>63.046898</td>\n",
       "      <td>100.793102</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         target                                                            \n",
       "           0.50                    0.90                    0.95            \n",
       "          lower       upper       lower       upper       lower       upper\n",
       "3    148.112984  179.007016  125.889924  201.230076  118.673336  208.446664\n",
       "197  172.995409  201.164591  152.732420  221.427580  146.152334  228.007666\n",
       "395  115.078302  138.421698   98.286654  155.213346   92.833832  160.666168\n",
       "78   162.962277  189.057723  144.190993  207.829007  138.095315  213.924685\n",
       "149  132.063311  155.816689  114.976752  172.903248  109.428161  178.451839\n",
       "..          ...         ...         ...         ...         ...         ...\n",
       "393   94.468654  120.131346   76.008663  138.591337   70.014072  144.585928\n",
       "316  216.443691  238.596309  200.508608  254.531392  195.333942  259.706058\n",
       "383   94.264854  116.395146   78.345830  132.314170   73.176378  137.483622\n",
       "122  261.935566  288.344434  242.938827  307.341173  236.769935  313.510065\n",
       "82    75.425129   88.414871   66.081193   97.758807   63.046898  100.793102\n",
       "\n",
       "[111 rows x 6 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coverage = [0.5, 0.9, 0.95]\n",
    "y_pred_ints = reg_proba.predict_interval(X_new, coverage=coverage)\n",
    "y_pred_ints"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`predict_interval` output spec:\n",
    "\n",
    "`pandas.DataFrame`\\\n",
    "Row index is as for `X_new`\\\n",
    "Column has multi-index:\\\n",
    "1st level = variable names from `y` in fit\\\n",
    "2nd level = coverage fractions in `coverage`\\\n",
    "3rd level = string `\"lower\"` or `\"upper\"`\n",
    "\n",
    "Entries = interval prediction of lower/upper interval at nominal coverage in 2nd lvl, for var in 1st lvl, for data index in row"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.1.3 quantile predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "quantile prediction `y_pred_quantiles` is a `pd.DataFrame`:\n",
    "\n",
    "* rows are the same as `X_new`\n",
    "* columns indicate variables, quantile points\n",
    "\n",
    "\"we predict the 5%, 50%, 95% quantile points for the row to be here\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"3\" halign=\"left\">target</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>0.05</th>\n",
       "      <th>0.50</th>\n",
       "      <th>0.95</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>125.889924</td>\n",
       "      <td>163.56</td>\n",
       "      <td>201.230076</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>152.732420</td>\n",
       "      <td>187.08</td>\n",
       "      <td>221.427580</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>98.286654</td>\n",
       "      <td>126.75</td>\n",
       "      <td>155.213346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>144.190993</td>\n",
       "      <td>176.01</td>\n",
       "      <td>207.829007</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>114.976752</td>\n",
       "      <td>143.94</td>\n",
       "      <td>172.903248</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         target                    \n",
       "           0.05    0.50        0.95\n",
       "3    125.889924  163.56  201.230076\n",
       "197  152.732420  187.08  221.427580\n",
       "395   98.286654  126.75  155.213346\n",
       "78   144.190993  176.01  207.829007\n",
       "149  114.976752  143.94  172.903248"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_quantiles = reg_proba.predict_quantiles(X_new, alpha=[0.05, 0.5, 0.95])\n",
    "y_pred_quantiles.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "multiple quantiles:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"5\" halign=\"left\">target</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>0.10</th>\n",
       "      <th>0.25</th>\n",
       "      <th>0.50</th>\n",
       "      <th>0.75</th>\n",
       "      <th>0.90</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>134.210188</td>\n",
       "      <td>148.112984</td>\n",
       "      <td>163.56</td>\n",
       "      <td>179.007016</td>\n",
       "      <td>192.909812</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>160.318837</td>\n",
       "      <td>172.995409</td>\n",
       "      <td>187.08</td>\n",
       "      <td>201.164591</td>\n",
       "      <td>213.841163</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>104.573410</td>\n",
       "      <td>115.078302</td>\n",
       "      <td>126.75</td>\n",
       "      <td>138.421698</td>\n",
       "      <td>148.926590</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>151.218920</td>\n",
       "      <td>162.962277</td>\n",
       "      <td>176.01</td>\n",
       "      <td>189.057723</td>\n",
       "      <td>200.801080</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>121.373921</td>\n",
       "      <td>132.063311</td>\n",
       "      <td>143.94</td>\n",
       "      <td>155.816689</td>\n",
       "      <td>166.506079</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>393</th>\n",
       "      <td>82.920042</td>\n",
       "      <td>94.468654</td>\n",
       "      <td>107.30</td>\n",
       "      <td>120.131346</td>\n",
       "      <td>131.679958</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>316</th>\n",
       "      <td>206.474668</td>\n",
       "      <td>216.443691</td>\n",
       "      <td>227.52</td>\n",
       "      <td>238.596309</td>\n",
       "      <td>248.565332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>383</th>\n",
       "      <td>84.305877</td>\n",
       "      <td>94.264854</td>\n",
       "      <td>105.33</td>\n",
       "      <td>116.395146</td>\n",
       "      <td>126.354123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>122</th>\n",
       "      <td>250.051163</td>\n",
       "      <td>261.935566</td>\n",
       "      <td>275.14</td>\n",
       "      <td>288.344434</td>\n",
       "      <td>300.228837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>82</th>\n",
       "      <td>69.579542</td>\n",
       "      <td>75.425129</td>\n",
       "      <td>81.92</td>\n",
       "      <td>88.414871</td>\n",
       "      <td>94.260458</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         target                                            \n",
       "           0.10        0.25    0.50        0.75        0.90\n",
       "3    134.210188  148.112984  163.56  179.007016  192.909812\n",
       "197  160.318837  172.995409  187.08  201.164591  213.841163\n",
       "395  104.573410  115.078302  126.75  138.421698  148.926590\n",
       "78   151.218920  162.962277  176.01  189.057723  200.801080\n",
       "149  121.373921  132.063311  143.94  155.816689  166.506079\n",
       "..          ...         ...     ...         ...         ...\n",
       "393   82.920042   94.468654  107.30  120.131346  131.679958\n",
       "316  206.474668  216.443691  227.52  238.596309  248.565332\n",
       "383   84.305877   94.264854  105.33  116.395146  126.354123\n",
       "122  250.051163  261.935566  275.14  288.344434  300.228837\n",
       "82    69.579542   75.425129   81.92   88.414871   94.260458\n",
       "\n",
       "[111 rows x 5 columns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "alpha = [0.1, 0.25, 0.5, 0.75, 0.9]\n",
    "y_pred_quantiles = reg_proba.predict_quantiles(X_new, alpha=alpha)\n",
    "y_pred_quantiles"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`predict_quantiles` output spec:\n",
    "\n",
    "`pandas.DataFrame`\\\n",
    "Row index is same as `X_new`\\\n",
    "Column has multi-index:\\\n",
    "1st level = variable names from `y` in `fit`\\\n",
    "2nd level = quantile points in `alpha`\n",
    "\n",
    "Entries = quantiles prediction at quantile point in 2nd lvl, for var in 1st lvl, for data index in row"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.1.4 mean and variance predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "mean and variance predictions `y_pred_mean`, `y_pred_var` are `pd.DataFrame`-s:\n",
    "\n",
    "* rows are the same as `X_new`\n",
    "* columns are the same as `X_new`\n",
    "\n",
    "entries are predictive mean and variance in row/column"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred_mean = reg_proba.predict(X_new)\n",
    "y_pred_var = reg_proba.predict_var(X_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3      163.56\n",
       "197    187.08\n",
       "395    126.75\n",
       "78     176.01\n",
       "149    143.94\n",
       "dtype: float64"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_mean.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>524.491525</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>436.051498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>299.445215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>374.212877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>310.055919</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         target\n",
       "3    524.491525\n",
       "197  436.051498\n",
       "395  299.445215\n",
       "78   374.212877\n",
       "149  310.055919"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_var.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "this is the same as taking the distribution prediction and taking mean/variance\n",
    "\n",
    "(for distribution objects that estimate these precisely)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>163.56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>187.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>126.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>176.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>143.94</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     target\n",
       "3    163.56\n",
       "197  187.08\n",
       "395  126.75\n",
       "78   176.01\n",
       "149  143.94"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_proba.mean().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>524.491525</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>197</th>\n",
       "      <td>436.051498</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>395</th>\n",
       "      <td>299.445215</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>374.212877</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>310.055919</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         target\n",
       "3    524.491525\n",
       "197  436.051498\n",
       "395  299.445215\n",
       "78   374.212877\n",
       "149  310.055919"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_proba.var().head()"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.2 simple evaluation workflow for probabilistic predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "for simple evaluation:\n",
    "\n",
    "1. split the data into train/test set\n",
    "2. make predictions of either type for test features\n",
    "3. compute metric on test set, comparing test predictions to held out test labels\n",
    "\n",
    "Note:\n",
    "\n",
    "* metrics will compare tabular ground truth to probabilistic prediction\n",
    "* the metric needs to be of a compatible type, e.g., for distribution predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "27.880267965820437"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.datasets import load_diabetes\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from skpro.metrics import CRPS\n",
    "from skpro.regression.residual import ResidualDouble\n",
    "\n",
    "# step 1: data specification\n",
    "X, y = load_diabetes(return_X_y=True, as_frame=True)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y)\n",
    "\n",
    "# step 2: specifying the regressor\n",
    "# example - linear regression for mean prediction\n",
    "# random forest for variance prediction\n",
    "reg_mean = LinearRegression()\n",
    "reg_resid = RandomForestRegressor()\n",
    "reg_proba = ResidualDouble(reg_mean, reg_resid)\n",
    "\n",
    "# step 3: fitting the model to training data\n",
    "reg_proba.fit(X_train, y_train)\n",
    "\n",
    "# step 4: predicting labels on new data\n",
    "y_pred_proba = reg_proba.predict_proba(X_test)\n",
    "\n",
    "# step 5: specifying evaluation metric\n",
    "metric = CRPS()\n",
    "\n",
    "# step 6: evaluate metric, compare predictions to actuals\n",
    "metric(y_test, y_pred_proba)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "how do we know that metric is of right type? Via `scitype:y_pred` tag"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'estimator_type': 'estimator',\n",
       " 'authors': 'skpro developers',\n",
       " 'maintainers': 'skpro developers',\n",
       " 'object_type': ['metric', 'metric_distr'],\n",
       " 'reserved_params': ['multioutput', 'score_average'],\n",
       " 'scitype:y_pred': 'pred_proba',\n",
       " 'lower_is_better': True}"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metric.get_tags()\n",
    "# scitype:y_pred is pred_proba - for proba predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "how do we find metrics for a prediction type?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>object</th>\n",
       "      <th>scitype:y_pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CRPS</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.CRPS'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ConcordanceHarrell</td>\n",
       "      <td>&lt;class 'skpro.metrics.survival._c_harrell.Conc...</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ConstraintViolation</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.ConstraintViola...</td>\n",
       "      <td>pred_interval</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>EmpiricalCoverage</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.EmpiricalCovera...</td>\n",
       "      <td>pred_interval</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LinearizedLogLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.LinearizedLogLo...</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>LogLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.LogLoss'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>PinballLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.PinballLoss'&gt;</td>\n",
       "      <td>pred_quantiles</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>SPLL</td>\n",
       "      <td>&lt;class 'skpro.metrics.survival._spll.SPLL'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>SquaredDistrLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.SquaredDistrLoss'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  name                                             object  \\\n",
       "0                 CRPS              <class 'skpro.metrics._classes.CRPS'>   \n",
       "1   ConcordanceHarrell  <class 'skpro.metrics.survival._c_harrell.Conc...   \n",
       "2  ConstraintViolation  <class 'skpro.metrics._classes.ConstraintViola...   \n",
       "3    EmpiricalCoverage  <class 'skpro.metrics._classes.EmpiricalCovera...   \n",
       "4    LinearizedLogLoss  <class 'skpro.metrics._classes.LinearizedLogLo...   \n",
       "5              LogLoss           <class 'skpro.metrics._classes.LogLoss'>   \n",
       "6          PinballLoss       <class 'skpro.metrics._classes.PinballLoss'>   \n",
       "7                 SPLL        <class 'skpro.metrics.survival._spll.SPLL'>   \n",
       "8     SquaredDistrLoss  <class 'skpro.metrics._classes.SquaredDistrLoss'>   \n",
       "\n",
       "   scitype:y_pred  \n",
       "0      pred_proba  \n",
       "1      pred_proba  \n",
       "2   pred_interval  \n",
       "3   pred_interval  \n",
       "4      pred_proba  \n",
       "5      pred_proba  \n",
       "6  pred_quantiles  \n",
       "7      pred_proba  \n",
       "8      pred_proba  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from skpro.registry import all_objects\n",
    "\n",
    "all_objects(\"metric\", as_dataframe=True, return_tags=\"scitype:y_pred\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "extra note: quantile metrics can be applied to interval predictions as well\n",
    "\n",
    "more details on metrics below"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 1.3 diagnostic visualisations"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "some useful diagnostic visualisations: variants of crossplots for probabilistic predictions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A. crossplot ground truth vs prediction intervals.\n",
    "\n",
    "Works with both proba and interval predictions.\n",
    "\n",
    "What to look for: intervals shouhld cut through the x = y line (green points)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Correct label $y_i$', ylabel='Prediction interval $\\\\widehat{y}_i$'>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from skpro.utils.plotting import plot_crossplot_interval\n",
    "\n",
    "plot_crossplot_interval(y_test, y_pred_proba, coverage=0.9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Correct label $y_i$', ylabel='Prediction interval $\\\\widehat{y}_i$'>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from skpro.utils.plotting import plot_crossplot_interval\n",
    "\n",
    "y_pred_interval = reg_proba.predict_interval(X_test, coverage=0.9)\n",
    "plot_crossplot_interval(y_test, y_pred_interval)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "B. crossplot residuals vs predictive standard deviation\n",
    "\n",
    "Works with both proba and variance predictions.\n",
    "\n",
    "What to look for: should be close to a line, high linear correlation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Absolute errors $|y_i - \\\\widehat{y}_i|$', ylabel='Predictive standard deviation of $\\\\widehat{y}_i$'>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from skpro.utils.plotting import plot_crossplot_std\n",
    "\n",
    "plot_crossplot_std(y_test, y_pred_proba)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Absolute errors $|y_i - \\\\widehat{y}_i|$', ylabel='Predictive standard deviation of $\\\\widehat{y}_i$'>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from skpro.utils.plotting import plot_crossplot_std\n",
    "\n",
    "y_pred_var = reg_proba.predict_var(X_test)\n",
    "plot_crossplot_std(y_test, y_pred_var)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "C. crossplot ground truth vs loss values\n",
    "\n",
    "Loss and prediction type should agree.\n",
    "\n",
    "What to look for: association between accuracy and ground truth value\n",
    "\n",
    "Diagnostic of which values we can predict more accurately,\n",
    "\n",
    "e.g., to inform modelling or identify unusual outliers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: title={'center': 'mean CRPS: 27.88 +/- 2.06 sterr of mean'}, xlabel='Correct label $y_i$', ylabel='CRPS($y_i$, $\\\\widehat{y}_i$)'>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from skpro.utils.plotting import plot_crossplot_loss\n",
    "\n",
    "crps_metric = CRPS()\n",
    "plot_crossplot_loss(y_test, y_pred_proba, crps_metric)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.4 `skpro` objects - `scikit-base` interface, searching for regressors and metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* `skpro` objects - `skbase` interface points `get_tags`, `get_params`/`set_params`\n",
    "* searching estimators and metrics via `all_objects`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.4.1 primer on `skpro` object interface <a class=\"anchor\" id=\"section1_3_1\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "metrics and estimators are first-class citizens in `skpro`, with a `scikit-base` compatible interface"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# example object 1: CRPS metric\n",
    "from skpro.metrics import CRPS\n",
    "\n",
    "crps_metric = CRPS()\n",
    "\n",
    "# example object 2: ResidualDouble regressor\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "from skpro.regression.residual import ResidualDouble\n",
    "\n",
    "reg_mean = LinearRegression()\n",
    "reg_resid = RandomForestRegressor()\n",
    "reg_proba = ResidualDouble(reg_mean, reg_resid)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "e.g., all have `get_tags` interface"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'estimator_type': 'estimator',\n",
       " 'authors': 'skpro developers',\n",
       " 'maintainers': 'skpro developers',\n",
       " 'object_type': ['metric', 'metric_distr'],\n",
       " 'reserved_params': ['multioutput', 'score_average'],\n",
       " 'scitype:y_pred': 'pred_proba',\n",
       " 'lower_is_better': True}"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "crps_metric.get_tags()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'estimator_type': 'regressor_proba',\n",
       " 'authors': 'skpro developers',\n",
       " 'maintainers': 'skpro developers',\n",
       " 'object_type': 'regressor_proba',\n",
       " 'capability:survival': False,\n",
       " 'capability:multioutput': False,\n",
       " 'capability:missing': True,\n",
       " 'X_inner_mtype': 'pd_DataFrame_Table',\n",
       " 'y_inner_mtype': 'pd_DataFrame_Table',\n",
       " 'C_inner_mtype': 'pd_DataFrame_Table'}"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reg_proba.get_tags()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "the tag `object_type` indicates the type of object, e.g., metric or proba regressor"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "all objects also have the `get_params`/`set_params` interface known from `scikit-learn`\n",
    "\n",
    "= reading or setting hyper-parameters\n",
    "\n",
    "`get_params` returns `dict` `{paramname: paramvalue}`; `set_params` writes it"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'multioutput': 'uniform_average', 'multivariate': False}"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "crps_metric.get_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "composite objects have the nested param interface, keys `componentname__paramname`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 {color: black;background-color: white;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 pre{padding: 0;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-toggleable {background-color: white;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-estimator:hover {background-color: #d4ebff;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-item {z-index: 1;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-parallel-item:only-child::after {width: 0;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-b580b327-619b-4f66-ad9b-1fa8646c3480 div.sk-text-repr-fallback {display: none;}</style><div id='sk-b580b327-619b-4f66-ad9b-1fa8646c3480' class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>ResidualDouble(estimator=LinearRegression(),\n",
       "               estimator_resid=RandomForestRegressor())</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class='sk-label-container'><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('635812d9-c1c9-47af-b354-9d2cba5e507d') type=\"checkbox\" ><label for=UUID('635812d9-c1c9-47af-b354-9d2cba5e507d') class='sk-toggleable__label sk-toggleable__label-arrow'>ResidualDouble</label><div class=\"sk-toggleable__content\"><pre>ResidualDouble(estimator=LinearRegression(),\n",
       "               estimator_resid=RandomForestRegressor())</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class='sk-item'><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('ff9f5c88-9d5c-4bef-98d9-b100e8db76dc') type=\"checkbox\" ><label for=UUID('ff9f5c88-9d5c-4bef-98d9-b100e8db76dc') class='sk-toggleable__label sk-toggleable__label-arrow'>LinearRegression</label><div class=\"sk-toggleable__content\"><pre>LinearRegression()</pre></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class='sk-item'><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('050e6606-48cf-419b-b867-0bd7af5c2b62') type=\"checkbox\" ><label for=UUID('050e6606-48cf-419b-b867-0bd7af5c2b62') class='sk-toggleable__label sk-toggleable__label-arrow'>RandomForestRegressor</label><div class=\"sk-toggleable__content\"><pre>RandomForestRegressor()</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "ResidualDouble(estimator=LinearRegression(),\n",
       "               estimator_resid=RandomForestRegressor())"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# note that reg_proba has components LinearRegression and RandomForestaregressor\n",
    "# each with their own parameters\n",
    "reg_proba"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "so `reg_proba` will have parameters coming from itself and either component:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'cv': None,\n",
       " 'distr_loc_scale_name': None,\n",
       " 'distr_params': None,\n",
       " 'distr_type': 'Normal',\n",
       " 'estimator': LinearRegression(),\n",
       " 'estimator_resid': RandomForestRegressor(),\n",
       " 'min_scale': 1e-10,\n",
       " 'residual_trafo': 'absolute',\n",
       " 'use_y_pred': False,\n",
       " 'estimator__copy_X': True,\n",
       " 'estimator__fit_intercept': True,\n",
       " 'estimator__n_jobs': None,\n",
       " 'estimator__positive': False,\n",
       " 'estimator_resid__bootstrap': True,\n",
       " 'estimator_resid__ccp_alpha': 0.0,\n",
       " 'estimator_resid__criterion': 'squared_error',\n",
       " 'estimator_resid__max_depth': None,\n",
       " 'estimator_resid__max_features': 1.0,\n",
       " 'estimator_resid__max_leaf_nodes': None,\n",
       " 'estimator_resid__max_samples': None,\n",
       " 'estimator_resid__min_impurity_decrease': 0.0,\n",
       " 'estimator_resid__min_samples_leaf': 1,\n",
       " 'estimator_resid__min_samples_split': 2,\n",
       " 'estimator_resid__min_weight_fraction_leaf': 0.0,\n",
       " 'estimator_resid__n_estimators': 100,\n",
       " 'estimator_resid__n_jobs': None,\n",
       " 'estimator_resid__oob_score': False,\n",
       " 'estimator_resid__random_state': None,\n",
       " 'estimator_resid__verbose': 0,\n",
       " 'estimator_resid__warm_start': False}"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "reg_proba.get_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "further common interface points are `get_config`, `set_config`, and `get_fitted_params` (only fittable estimators)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 1.4.2 searching for regressors and metrics <a class=\"anchor\" id=\"section1_3_2\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "as first-class citizens, all objects in `skpro` are indexed via the `registry` utility `all_objects`.\n",
    "\n",
    "To find probabilistic supervised regressors, use `all_objects` with the type `regressor_proba`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>object</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>AFTFisk</td>\n",
       "      <td>&lt;class 'skpro.survival.aft._aft_lifelines_fisk...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>AFTLogNormal</td>\n",
       "      <td>&lt;class 'skpro.survival.aft._aft_lifelines_logn...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AFTWeibull</td>\n",
       "      <td>&lt;class 'skpro.survival.aft._aft_lifelines_weib...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>ARDRegression</td>\n",
       "      <td>&lt;class 'skpro.regression.linear._sklearn.ARDRe...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>AalenAdditive</td>\n",
       "      <td>&lt;class 'skpro.survival.additive._aalen_lifelin...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "            name                                             object\n",
       "0        AFTFisk  <class 'skpro.survival.aft._aft_lifelines_fisk...\n",
       "1   AFTLogNormal  <class 'skpro.survival.aft._aft_lifelines_logn...\n",
       "2     AFTWeibull  <class 'skpro.survival.aft._aft_lifelines_weib...\n",
       "3  ARDRegression  <class 'skpro.regression.linear._sklearn.ARDRe...\n",
       "4  AalenAdditive  <class 'skpro.survival.additive._aalen_lifelin..."
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from skpro.registry import all_objects\n",
    "\n",
    "all_objects(\"regressor_proba\", as_dataframe=True).head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "a full list can also be found in the online API reference."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "for metrics, as seen above:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>object</th>\n",
       "      <th>scitype:y_pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CRPS</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.CRPS'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ConcordanceHarrell</td>\n",
       "      <td>&lt;class 'skpro.metrics.survival._c_harrell.Conc...</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ConstraintViolation</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.ConstraintViola...</td>\n",
       "      <td>pred_interval</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>EmpiricalCoverage</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.EmpiricalCovera...</td>\n",
       "      <td>pred_interval</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LinearizedLogLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.LinearizedLogLo...</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>LogLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.LogLoss'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>PinballLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.PinballLoss'&gt;</td>\n",
       "      <td>pred_quantiles</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>SPLL</td>\n",
       "      <td>&lt;class 'skpro.metrics.survival._spll.SPLL'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>SquaredDistrLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.SquaredDistrLoss'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  name                                             object  \\\n",
       "0                 CRPS              <class 'skpro.metrics._classes.CRPS'>   \n",
       "1   ConcordanceHarrell  <class 'skpro.metrics.survival._c_harrell.Conc...   \n",
       "2  ConstraintViolation  <class 'skpro.metrics._classes.ConstraintViola...   \n",
       "3    EmpiricalCoverage  <class 'skpro.metrics._classes.EmpiricalCovera...   \n",
       "4    LinearizedLogLoss  <class 'skpro.metrics._classes.LinearizedLogLo...   \n",
       "5              LogLoss           <class 'skpro.metrics._classes.LogLoss'>   \n",
       "6          PinballLoss       <class 'skpro.metrics._classes.PinballLoss'>   \n",
       "7                 SPLL        <class 'skpro.metrics.survival._spll.SPLL'>   \n",
       "8     SquaredDistrLoss  <class 'skpro.metrics._classes.SquaredDistrLoss'>   \n",
       "\n",
       "   scitype:y_pred  \n",
       "0      pred_proba  \n",
       "1      pred_proba  \n",
       "2   pred_interval  \n",
       "3   pred_interval  \n",
       "4      pred_proba  \n",
       "5      pred_proba  \n",
       "6  pred_quantiles  \n",
       "7      pred_proba  \n",
       "8      pred_proba  "
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from skpro.registry import all_objects\n",
    "\n",
    "all_objects(\"metric\", as_dataframe=True, return_tags=\"scitype:y_pred\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "all tags can be printed by the `all_tags` utility:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>scitype</th>\n",
       "      <th>type</th>\n",
       "      <th>description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>capability:survival</td>\n",
       "      <td>metric</td>\n",
       "      <td>bool</td>\n",
       "      <td>whether metric uses censoring information, for...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>lower_is_better</td>\n",
       "      <td>metric</td>\n",
       "      <td>bool</td>\n",
       "      <td>whether lower (True) or higher (False) is better</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>scitype:y_pred</td>\n",
       "      <td>metric</td>\n",
       "      <td>str</td>\n",
       "      <td>expected input type for y_pred in performance ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  name scitype  type  \\\n",
       "0  capability:survival  metric  bool   \n",
       "1      lower_is_better  metric  bool   \n",
       "2       scitype:y_pred  metric   str   \n",
       "\n",
       "                                         description  \n",
       "0  whether metric uses censoring information, for...  \n",
       "1   whether lower (True) or higher (False) is better  \n",
       "2  expected input type for y_pred in performance ...  "
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# all tags applicable to metrics\n",
    "from skpro.registry import all_tags\n",
    "\n",
    "all_tags(\"metric\", as_dataframe=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>scitype</th>\n",
       "      <th>type</th>\n",
       "      <th>description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>C_inner_mtype</td>\n",
       "      <td>regressor_proba</td>\n",
       "      <td>(list, str)</td>\n",
       "      <td>which machine type(s) is the internal _fit/_pr...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>X_inner_mtype</td>\n",
       "      <td>regressor_proba</td>\n",
       "      <td>(list, str)</td>\n",
       "      <td>which machine type(s) is the internal _fit/_pr...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>capability:missing</td>\n",
       "      <td>regressor_proba</td>\n",
       "      <td>bool</td>\n",
       "      <td>whether estimator supports missing values</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>capability:multioutput</td>\n",
       "      <td>regressor_proba</td>\n",
       "      <td>bool</td>\n",
       "      <td>whether estimator supports multioutput regression</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>capability:survival</td>\n",
       "      <td>regressor_proba</td>\n",
       "      <td>bool</td>\n",
       "      <td>whether estimator can use censoring information...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>y_inner_mtype</td>\n",
       "      <td>regressor_proba</td>\n",
       "      <td>(list, str)</td>\n",
       "      <td>which machine type(s) is the internal _fit/_pr...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                     name          scitype         type  \\\n",
       "0           C_inner_mtype  regressor_proba  (list, str)   \n",
       "1           X_inner_mtype  regressor_proba  (list, str)   \n",
       "2      capability:missing  regressor_proba         bool   \n",
       "3  capability:multioutput  regressor_proba         bool   \n",
       "4     capability:survival  regressor_proba         bool   \n",
       "5           y_inner_mtype  regressor_proba  (list, str)   \n",
       "\n",
       "                                         description  \n",
       "0  which machine type(s) is the internal _fit/_pr...  \n",
       "1  which machine type(s) is the internal _fit/_pr...  \n",
       "2          whether estimator supports missing values  \n",
       "3  whether estimator supports multioutput regression  \n",
       "4  whether estimator can use censoring information...  \n",
       "5  which machine type(s) is the internal _fit/_pr...  "
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# all tags applicable to probabilistic regressors\n",
    "from skpro.registry import all_tags\n",
    "\n",
    "all_tags(\"regressor_proba\", as_dataframe=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "filtering in search can be done with the `filter_tags` argument in `all_objects`, see docstring:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>object</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CRPS</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.CRPS'&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ConcordanceHarrell</td>\n",
       "      <td>&lt;class 'skpro.metrics.survival._c_harrell.Conc...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>LinearizedLogLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.LinearizedLogLo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>LogLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.LogLoss'&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SPLL</td>\n",
       "      <td>&lt;class 'skpro.metrics.survival._spll.SPLL'&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>SquaredDistrLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.SquaredDistrLoss'&gt;</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 name                                             object\n",
       "0                CRPS              <class 'skpro.metrics._classes.CRPS'>\n",
       "1  ConcordanceHarrell  <class 'skpro.metrics.survival._c_harrell.Conc...\n",
       "2   LinearizedLogLoss  <class 'skpro.metrics._classes.LinearizedLogLo...\n",
       "3             LogLoss           <class 'skpro.metrics._classes.LogLoss'>\n",
       "4                SPLL        <class 'skpro.metrics.survival._spll.SPLL'>\n",
       "5    SquaredDistrLoss  <class 'skpro.metrics._classes.SquaredDistrLoss'>"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from skpro.registry import all_objects\n",
    "\n",
    "# \"retrieve all genuinely probabilistic loss functions\"\n",
    "all_objects(\"metric\", as_dataframe=True, filter_tags={\"scitype:y_pred\": \"pred_proba\"})"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2. Prediction types, metrics, benchmarking <a class=\"anchor\" id=\"chapter2\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This section gives more details on:\n",
    "\n",
    "* different prediction types, including a methodological primer\n",
    "* the API of metrics to compare probabilistic predictions to non-probabilistic actuals\n",
    "* utilities for batch benchmarking of estimators and metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.1 Probabilistic predictions - methodological primer <a class=\"anchor\" id=\"section2_1\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**readers familiar with, or less interested in theory, may like to skip section 2.1**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In supervised learning - probabilistic or not:\n",
    "\n",
    "* we fit estimator to i.i.d samples $(X_1, Y_1), \\dots, (X_N, Y_N) \\sim (X_*, Y_*)$\n",
    "* and want to predict $y$ given $x$ accurately, for $(x, y) \\sim (X_*, Y_*)$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Let $y$ be the (true) value, for an observed feature $x$\n",
    "\n",
    "(we consider $y$ a random variable)\n",
    "\n",
    "| Name | param | prediction/estimate of | `skpro` |\n",
    "| ---- | ----- | ---------------------- | -------- |\n",
    "| point prediction | | conditional expectation $\\mathbb{E}[y\\|x]$ | `predict` |\n",
    "| variance prediction | | conditional variance $Var[y\\|x]$ | `predict_var` |\n",
    "| quantile prediction | $\\alpha\\in (0,1)$ | $\\alpha$-quantile of $y\\|x$ | `predict_quantiles` |\n",
    "| interval prediction | $c\\in (0,1)$| $[a,b]$ s.t. $P(a\\le y \\le b\\| x) = c$ | `predict_interval` |\n",
    "| distribution prediction | | the law/distribution of $y\\|x$ | `predict_proba` |"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##### More formal details & intuition:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "let's consider the toy example again"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_diabetes\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X, y = load_diabetes(return_X_y=True, as_frame=True)\n",
    "X_train, X_new, y_train, _ = train_test_split(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_diabetes\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "from skpro.regression.residual import ResidualDouble\n",
    "\n",
    "X, y = load_diabetes(return_X_y=True, as_frame=True)\n",
    "X_train, X_new, y_train, _ = train_test_split(X, y)\n",
    "\n",
    "\n",
    "reg_mean = RandomForestRegressor()\n",
    "reg_proba = ResidualDouble(reg_mean)\n",
    "\n",
    "reg_proba.fit(X_train, y_train)\n",
    "y_pred_proba = reg_proba.predict_proba(X_new)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* a **\"point prediction\"** is a prediction/estimate of the conditional expectation $\\mathbb{E}[y|x]$.\\\n",
    " **Intuition**: \"out of many repetitions/worlds, this value is the arithmetic average of all observations\"."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>138.470645</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>241</th>\n",
       "      <td>129.681224</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>133.812624</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>432</th>\n",
       "      <td>260.395023</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>73.546588</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         target\n",
       "102  138.470645\n",
       "241  129.681224\n",
       "282  133.812624\n",
       "432  260.395023\n",
       "33    73.546588"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# if y_pred_proba were *true*, here's how many repetitions would look like:\n",
    "\n",
    "# repeating this line is \"one repetition\"\n",
    "y_pred_proba.sample().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">0</th>\n",
       "      <th>102</th>\n",
       "      <td>146.649904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>241</th>\n",
       "      <td>141.674200</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>124.885507</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>432</th>\n",
       "      <td>226.703394</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>104.064224</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">99</th>\n",
       "      <th>220</th>\n",
       "      <td>38.277940</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>330</th>\n",
       "      <td>201.071314</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>213</th>\n",
       "      <td>96.107157</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>42.479152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>115</th>\n",
       "      <td>215.560492</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>11100 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "            target\n",
       "0  102  146.649904\n",
       "   241  141.674200\n",
       "   282  124.885507\n",
       "   432  226.703394\n",
       "   33   104.064224\n",
       "...            ...\n",
       "99 220   38.277940\n",
       "   330  201.071314\n",
       "   213   96.107157\n",
       "   6     42.479152\n",
       "   115  215.560492\n",
       "\n",
       "[11100 rows x 1 columns]"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "many_samples = y_pred_proba.sample(100)\n",
    "many_samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>149.718055</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>241</th>\n",
       "      <td>110.814113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>132.969532</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>432</th>\n",
       "      <td>230.375040</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>100.993639</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         target\n",
       "102  149.718055\n",
       "241  110.814113\n",
       "282  132.969532\n",
       "432  230.375040\n",
       "33   100.993639"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# \"doing many times and taking the mean\" -> usual point prediction\n",
    "mean_prediction = many_samples.groupby(level=1, sort=False).mean()\n",
    "mean_prediction.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>150.29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>241</th>\n",
       "      <td>113.42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>132.04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>432</th>\n",
       "      <td>231.15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>102.79</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     target\n",
       "102  150.29\n",
       "241  113.42\n",
       "282  132.04\n",
       "432  231.15\n",
       "33   102.79"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# if we would do this infinity times instead of 100:\n",
    "y_pred_proba.mean().head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* a **\"variance prediction\"** is a prediction/estimate of the conditional expectation $Var[y|x]$.\\\n",
    " **Intuition:** \"out of many repetitions/worlds, this value is the average squared distance of the observation to the perfect point prediction\".\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>326.063122</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>241</th>\n",
       "      <td>272.877878</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>288.577677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>432</th>\n",
       "      <td>280.419597</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>278.622982</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         target\n",
       "102  326.063122\n",
       "241  272.877878\n",
       "282  288.577677\n",
       "432  280.419597\n",
       "33   278.622982"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# same as above - take many samples, and then compute element-wise statistics\n",
    "var_prediction = many_samples.groupby(level=1, sort=False).var()\n",
    "var_prediction.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>102</th>\n",
       "      <td>273.410721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>241</th>\n",
       "      <td>273.410721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>282</th>\n",
       "      <td>273.410721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>432</th>\n",
       "      <td>273.410721</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>273.410721</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         target\n",
       "102  273.410721\n",
       "241  273.410721\n",
       "282  273.410721\n",
       "432  273.410721\n",
       "33   273.410721"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# e.g., predict_var should give the same result as infinite large sample's variance\n",
    "y_pred_proba.var().head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* a **\"quantile prediction\"**, at quantile point $\\alpha\\in (0,1)$ is a prediction/estimate of the $\\alpha$-quantile of $y'|y$, i.e., of $F^{-1}_{y|x}(\\alpha)$, where $F^{-1}$ is the (generalized) inverse cdf = quantile function of the random variable y|x.\\\n",
    " **Intuition**: \"out of many repetitions/worlds, a fraction of exactly $\\alpha$ will have equal or smaller than this value.\"\n",
    "* an **\"interval prediction\"** or \"predictive interval\" with (symmetric) coverage $c\\in (0,1)$ is a prediction/estimate pair of lower bound $a$ and upper bound $b$ such that $P(a\\le y \\le b| x) = c$ and $P(y \\gneq b| x) = P(y \\lneq a| x) = (1 - c) /2$.\\\n",
    " **Intuition**: \"out of many repetitions/worlds, a fraction of exactly $c$ will be contained in the interval $[a,b]$, and being above is equally likely as being below\"."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "(similar - exercise left to the reader)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* a **\"distribution prediction\"** or \"full probabilistic prediction\" is a prediction/estimate of the distribution of $y|x$, e.g., \"it's a normal distribution with mean 42 and variance 1\".\\\n",
    "**Intuition**: exhaustive description of the generating mechanism of many repetitions/worlds."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "note: the true distribution is unknown, and not accessible easily!\n",
    "\n",
    "`y_pred_proba` is a distribution, but in general not equal to the true one!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "that is, there are:\n",
    "\n",
    "* *true* distribution `y_pred_proba_true` - unknown and unknowable but estimable\n",
    "* `y_pred_proba` - our guess at `y_pred_proba_true`\n",
    "* the actual data `y_true` is *one* `y_pred_proba_true.sample()`\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* `predict` produces guess of `y_pred_proba_true.mean()`\n",
    "* `predict_var` produces guess of `y_pred_proba_true.var()`\n",
    "* `predict_quantiles([0.05, 0.5, 0.95])` produces guess of `y_pred_proba_true.quantiles([0.05, 0.5, 0.95])`\n",
    "* `predict_proba` produces guess of `y_pred_proba_true`\n",
    "\n",
    "the guesses are algorithm specific, and some algorithms are more accurate than others, given data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.2 probabilistic metrics - details <a class=\"anchor\" id=\"section2_2\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "General usage pattern same as for `sklearn` metrics:\n",
    "\n",
    "1. get some actuals and predictions\n",
    "2. specify the metric - similar to estimator specs\n",
    "3. plug the actuals and predictions into metric to get metric values\n",
    "\n",
    "*but*: need to use dedicated metric for probabilistic predictions\n",
    "\n",
    "* ground truth: `y_true` samples\n",
    "* prediction e.g., `y_predict_proba`, `y_predict_interval`\n",
    "* so, match metric with type of prediction!\n",
    "    * `metric(y_true: 2D pd.DataFrame, y_pred: proba_prediction_type) -> float`"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Recall methods available for all probabilistic regressors:\n",
    "\n",
    "- `predict_interval` produces interval predictions.\n",
    "  Argument `coverage` (nominal interval coverage) must be provided.\n",
    "- `predict_quantiles` produces quantile predictions.\n",
    "  Argument `alpha` (quantile values) must be provided.\n",
    "- `predict_var` produces variance predictions. Same args as `predict`.\n",
    "- `predict_proba` produces full distributional predictions. Same args as `predict`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "| Name | param | prediction/estimate of | `skpro` |\n",
    "| ---- | ----- | ---------------------- | -------- |\n",
    "| point prediction | | conditional expectation $\\mathbb{E}[y\\|x]$ | `predict` |\n",
    "| variance prediction | | conditional variance $Var[y\\|x]$ | `predict_var` |\n",
    "| quantile prediction | $\\alpha\\in (0,1)$ | $\\alpha$-quantile of $y\\|x$ | `predict_quantiles` |\n",
    "| interval prediction | $c\\in (0,1)$| $[a,b]$ s.t. $P(a\\le y \\le b\\| x) = c$ | `predict_interval` |\n",
    "| distribution prediction | | the law/distribution of $y\\|x$ | `predict_proba` |"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "let's produce some probabilistic predictions!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 1. get some actuals and predictions\n",
    "from sklearn.datasets import load_diabetes\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X, y = load_diabetes(return_X_y=True, as_frame=True)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y)\n",
    "# actuals = y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "\n",
    "from skpro.regression.residual import ResidualDouble\n",
    "\n",
    "reg_mean = RandomForestRegressor()\n",
    "reg_proba = ResidualDouble(reg_mean)\n",
    "\n",
    "reg_proba.fit(X_train, y_train)\n",
    "\n",
    "# use any of the probabilistic methods, we have seen this\n",
    "y_pred_int = reg_proba.predict_interval(X_test, coverage=0.95)\n",
    "y_pred_q = reg_proba.predict_quantiles(X_test, alpha=[0.05, 0.95])\n",
    "y_pred_proba = reg_proba.predict_proba(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "recall, all have their own output format:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">target</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">0.95</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>lower</th>\n",
       "      <th>upper</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>65.377963</td>\n",
       "      <td>132.482037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>38.007963</td>\n",
       "      <td>105.112037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>170</th>\n",
       "      <td>41.137963</td>\n",
       "      <td>108.242037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>76.017963</td>\n",
       "      <td>143.122037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>76.837963</td>\n",
       "      <td>143.942037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>135.807963</td>\n",
       "      <td>202.912037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>414</th>\n",
       "      <td>151.257963</td>\n",
       "      <td>218.362037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269</th>\n",
       "      <td>64.367963</td>\n",
       "      <td>131.472037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>279</th>\n",
       "      <td>64.107963</td>\n",
       "      <td>131.212037</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>232</th>\n",
       "      <td>124.657963</td>\n",
       "      <td>191.762037</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         target            \n",
       "           0.95            \n",
       "          lower       upper\n",
       "175   65.377963  132.482037\n",
       "93    38.007963  105.112037\n",
       "170   41.137963  108.242037\n",
       "105   76.017963  143.122037\n",
       "7     76.837963  143.942037\n",
       "..          ...         ...\n",
       "49   135.807963  202.912037\n",
       "414  151.257963  218.362037\n",
       "269   64.367963  131.472037\n",
       "279   64.107963  131.212037\n",
       "232  124.657963  191.762037\n",
       "\n",
       "[111 rows x 2 columns]"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_int  # lower/upper intervals"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">target</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th>0.05</th>\n",
       "      <th>0.95</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>70.772243</td>\n",
       "      <td>127.087757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>43.402243</td>\n",
       "      <td>99.717757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>170</th>\n",
       "      <td>46.532243</td>\n",
       "      <td>102.847757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>81.412243</td>\n",
       "      <td>137.727757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>82.232243</td>\n",
       "      <td>138.547757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>141.202243</td>\n",
       "      <td>197.517757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>414</th>\n",
       "      <td>156.652243</td>\n",
       "      <td>212.967757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269</th>\n",
       "      <td>69.762243</td>\n",
       "      <td>126.077757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>279</th>\n",
       "      <td>69.502243</td>\n",
       "      <td>125.817757</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>232</th>\n",
       "      <td>130.052243</td>\n",
       "      <td>186.367757</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         target            \n",
       "           0.05        0.95\n",
       "175   70.772243  127.087757\n",
       "93    43.402243   99.717757\n",
       "170   46.532243  102.847757\n",
       "105   81.412243  137.727757\n",
       "7     82.232243  138.547757\n",
       "..          ...         ...\n",
       "49   141.202243  197.517757\n",
       "414  156.652243  212.967757\n",
       "269   69.762243  126.077757\n",
       "279   69.502243  125.817757\n",
       "232  130.052243  186.367757\n",
       "\n",
       "[111 rows x 2 columns]"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_q  # quantiles"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 {color: black;background-color: white;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 pre{padding: 0;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-toggleable {background-color: white;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-estimator:hover {background-color: #d4ebff;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-item {z-index: 1;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-parallel-item:only-child::after {width: 0;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-b8f7296c-f305-48b6-8780-fc21d91f4535 div.sk-text-repr-fallback {display: none;}</style><div id='sk-b8f7296c-f305-48b6-8780-fc21d91f4535' class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Normal(columns=Index([&#x27;target&#x27;], dtype=&#x27;object&#x27;),\n",
       "       index=Index([175,  93, 170, 105,   7, 264, 133, 237, 274, 370,\n",
       "       ...\n",
       "       173, 253,  53, 212, 392,  49, 414, 269, 279, 232],\n",
       "      dtype=&#x27;int64&#x27;, length=111),\n",
       "       mu=array([[ 98.93],\n",
       "       [ 71.56],\n",
       "       [ 74.69],\n",
       "       [109.57],\n",
       "       [110.39],\n",
       "       [141.99],\n",
       "       [ 83.61],\n",
       "       [ 79.87],\n",
       "       [144.44],\n",
       "       [ 75.95],\n",
       "       [180.99],\n",
       "       [ 97.24],\n",
       "       [153.78],\n",
       "       [ 93.94],\n",
       "       [126.26],\n",
       "       [149.51],\n",
       "       [233.92],\n",
       "       [113.2 ],\n",
       "       [181.91],\n",
       "       [111.28],\n",
       "       [24...\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091]]))</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class='sk-item'><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('2937dacb-d73b-4ddc-854d-d8bf7066c517') type=\"checkbox\" checked><label for=UUID('2937dacb-d73b-4ddc-854d-d8bf7066c517') class='sk-toggleable__label sk-toggleable__label-arrow'>Normal</label><div class=\"sk-toggleable__content\"><pre>Normal(columns=Index([&#x27;target&#x27;], dtype=&#x27;object&#x27;),\n",
       "       index=Index([175,  93, 170, 105,   7, 264, 133, 237, 274, 370,\n",
       "       ...\n",
       "       173, 253,  53, 212, 392,  49, 414, 269, 279, 232],\n",
       "      dtype=&#x27;int64&#x27;, length=111),\n",
       "       mu=array([[ 98.93],\n",
       "       [ 71.56],\n",
       "       [ 74.69],\n",
       "       [109.57],\n",
       "       [110.39],\n",
       "       [141.99],\n",
       "       [ 83.61],\n",
       "       [ 79.87],\n",
       "       [144.44],\n",
       "       [ 75.95],\n",
       "       [180.99],\n",
       "       [ 97.24],\n",
       "       [153.78],\n",
       "       [ 93.94],\n",
       "       [126.26],\n",
       "       [149.51],\n",
       "       [233.92],\n",
       "       [113.2 ],\n",
       "       [181.91],\n",
       "       [111.28],\n",
       "       [24...\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091]]))</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "Normal(columns=Index(['target'], dtype='object'),\n",
       "       index=Index([175,  93, 170, 105,   7, 264, 133, 237, 274, 370,\n",
       "       ...\n",
       "       173, 253,  53, 212, 392,  49, 414, 269, 279, 232],\n",
       "      dtype='int64', length=111),\n",
       "       mu=array([[ 98.93],\n",
       "       [ 71.56],\n",
       "       [ 74.69],\n",
       "       [109.57],\n",
       "       [110.39],\n",
       "       [141.99],\n",
       "       [ 83.61],\n",
       "       [ 79.87],\n",
       "       [144.44],\n",
       "       [ 75.95],\n",
       "       [180.99],\n",
       "       [ 97.24],\n",
       "       [153.78],\n",
       "       [ 93.94],\n",
       "       [126.26],\n",
       "       [149.51],\n",
       "       [233.92],\n",
       "       [113.2 ],\n",
       "       [181.91],\n",
       "       [111.28],\n",
       "       [24...\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091],\n",
       "       [17.11870091]]))"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_proba  # sktime/skpro BaseDistribution"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "we now need to apply a suitable metric, `metric(y_test, y_pred)`\n",
    "\n",
    "IMPORTANT: sequence matters, `y_test` first; `y_pred` has very different type!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "38.547321051730464"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 2. specify metric\n",
    "# CRPS = continuous ranked probability score, for distribution predictions\n",
    "from skpro.metrics import CRPS\n",
    "\n",
    "crps = CRPS()\n",
    "\n",
    "# 3. evaluate metric\n",
    "crps(y_test, y_pred_proba)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "how do we find a metric that fits the prediction type?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "answer: metrics are tagged\n",
    "\n",
    "important tag: `scitype:y_pred`\n",
    "\n",
    "* `\"pred_proba\"` - distributional, can applied to distributions, `predict_proba` output\n",
    "* `\"pred_quantiles\"` - quantile forecast metric, can be applied to quantile predictions, interval predictions, distributional predictions\n",
    "    * applicable to `predict_quantiles`, `predict_interval`, `predict_proba` outputs\n",
    "* `\"pred_interval\"` - interval forecast metric, can be applied to interval predictions, distributional predictions\n",
    "    * applicable to `predict_interval`, `predict_proba` outputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'estimator_type': 'estimator',\n",
       " 'authors': 'skpro developers',\n",
       " 'maintainers': 'skpro developers',\n",
       " 'object_type': ['metric', 'metric_distr'],\n",
       " 'reserved_params': ['multioutput', 'score_average'],\n",
       " 'scitype:y_pred': 'pred_proba',\n",
       " 'lower_is_better': True}"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "crps.get_tags()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "listing metrics with the tag, filtering for probabilistic tags:\n",
    "\n",
    "(let's try to find a quantile prediction metric!)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>name</th>\n",
       "      <th>object</th>\n",
       "      <th>scitype:y_pred</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>CRPS</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.CRPS'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>ConcordanceHarrell</td>\n",
       "      <td>&lt;class 'skpro.metrics.survival._c_harrell.Conc...</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ConstraintViolation</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.ConstraintViola...</td>\n",
       "      <td>pred_interval</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>EmpiricalCoverage</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.EmpiricalCovera...</td>\n",
       "      <td>pred_interval</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LinearizedLogLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.LinearizedLogLo...</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>LogLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.LogLoss'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>PinballLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.PinballLoss'&gt;</td>\n",
       "      <td>pred_quantiles</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>SPLL</td>\n",
       "      <td>&lt;class 'skpro.metrics.survival._spll.SPLL'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>SquaredDistrLoss</td>\n",
       "      <td>&lt;class 'skpro.metrics._classes.SquaredDistrLoss'&gt;</td>\n",
       "      <td>pred_proba</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  name                                             object  \\\n",
       "0                 CRPS              <class 'skpro.metrics._classes.CRPS'>   \n",
       "1   ConcordanceHarrell  <class 'skpro.metrics.survival._c_harrell.Conc...   \n",
       "2  ConstraintViolation  <class 'skpro.metrics._classes.ConstraintViola...   \n",
       "3    EmpiricalCoverage  <class 'skpro.metrics._classes.EmpiricalCovera...   \n",
       "4    LinearizedLogLoss  <class 'skpro.metrics._classes.LinearizedLogLo...   \n",
       "5              LogLoss           <class 'skpro.metrics._classes.LogLoss'>   \n",
       "6          PinballLoss       <class 'skpro.metrics._classes.PinballLoss'>   \n",
       "7                 SPLL        <class 'skpro.metrics.survival._spll.SPLL'>   \n",
       "8     SquaredDistrLoss  <class 'skpro.metrics._classes.SquaredDistrLoss'>   \n",
       "\n",
       "   scitype:y_pred  \n",
       "0      pred_proba  \n",
       "1      pred_proba  \n",
       "2   pred_interval  \n",
       "3   pred_interval  \n",
       "4      pred_proba  \n",
       "5      pred_proba  \n",
       "6  pred_quantiles  \n",
       "7      pred_proba  \n",
       "8      pred_proba  "
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from skpro.registry import all_objects\n",
    "\n",
    "all_objects(\n",
    "    \"metric\",\n",
    "    as_dataframe=True,\n",
    "    return_tags=\"scitype:y_pred\",\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`PinballLoss` is a quantile forecast metric:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "13.124000434829115"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from skpro.metrics import PinballLoss\n",
    "\n",
    "pinball_loss = PinballLoss()\n",
    "\n",
    "pinball_loss(y_test, y_pred_q)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "... this is by default an average (grand average, float)\n",
    "\n",
    "* averages over samples in `y_pred` / `y_test` (rows)\n",
    "* averages over variables (columns)\n",
    "* average over `alpha` values, quantile points"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "what if we don't want these averages?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* variable (column) averaging is controlled by the `multioutput` arg.\n",
    "    * `\"raw_values\"` prevents averaging, `\"uniform_average\"` computes arithmetic mean.\n",
    "* quantile points (`alpha`) or coverage (`coverage`) is controlled by `score_average` arg\n",
    "* evaluation by row via the `evaluate_by_index` method\n",
    "    * can be useful for diagnostics or statistical tests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.05    10.994424\n",
       "0.95    15.253577\n",
       "Name: 0, dtype: float64"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Example 1: Pinball loss by quantile point\n",
    "loss_multi = PinballLoss(score_average=False)\n",
    "loss_multi(y_test, y_pred_q)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>175</th>\n",
       "      <td>19.135950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>93</th>\n",
       "      <td>15.962640</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>170</th>\n",
       "      <td>18.795242</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>105</th>\n",
       "      <td>46.916090</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>37.760777</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>18.500837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>414</th>\n",
       "      <td>44.159622</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>269</th>\n",
       "      <td>6.688994</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>279</th>\n",
       "      <td>92.681807</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>232</th>\n",
       "      <td>91.131807</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>111 rows × 1 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        target\n",
       "175  19.135950\n",
       "93   15.962640\n",
       "170  18.795242\n",
       "105  46.916090\n",
       "7    37.760777\n",
       "..         ...\n",
       "49   18.500837\n",
       "414  44.159622\n",
       "269   6.688994\n",
       "279  92.681807\n",
       "232  91.131807\n",
       "\n",
       "[111 rows x 1 columns]"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Example 2: CRPS by test sample index\n",
    "crps.evaluate_by_index(y_test, y_pred_proba)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Caveat: not every metric is an average over time points, e.g., RMSE\n",
    "\n",
    "In this case, `evaluate_by_index` computes jackknife pseudo-samples\n",
    "\n",
    "(for mean statistics, jackknife pseudo-samples are equal to individual samples)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 2.3 Benchmark evaluation of probabilistic regressors <a class=\"anchor\" id=\"section2_3\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "for quick evaluation and benchmarking,\n",
    "\n",
    "the `benchmarking.evaluate` utility can be used:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>test_CRPS</th>\n",
       "      <th>fit_time</th>\n",
       "      <th>pred_proba_time</th>\n",
       "      <th>len_y_train</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>31.783451</td>\n",
       "      <td>0.005134</td>\n",
       "      <td>0.002294</td>\n",
       "      <td>294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>33.574329</td>\n",
       "      <td>0.005496</td>\n",
       "      <td>0.002145</td>\n",
       "      <td>295</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>29.909655</td>\n",
       "      <td>0.004958</td>\n",
       "      <td>0.002201</td>\n",
       "      <td>295</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   test_CRPS  fit_time  pred_proba_time  len_y_train\n",
       "0  31.783451  0.005134         0.002294          294\n",
       "1  33.574329  0.005496         0.002145          295\n",
       "2  29.909655  0.004958         0.002201          295"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.datasets import load_diabetes\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.model_selection import KFold\n",
    "\n",
    "from skpro.benchmarking.evaluate import evaluate\n",
    "from skpro.metrics import CRPS\n",
    "from skpro.regression.residual import ResidualDouble\n",
    "\n",
    "# 1. specify dataset\n",
    "X, y = load_diabetes(return_X_y=True, as_frame=True)\n",
    "\n",
    "# 2. specify estimator\n",
    "estimator = ResidualDouble(LinearRegression())\n",
    "\n",
    "# 3. specify cross-validation schema\n",
    "cv = KFold(n_splits=3)\n",
    "\n",
    "# 4. specify evaluation metric\n",
    "crps = CRPS()\n",
    "\n",
    "# 5. evaluate - run the benchmark\n",
    "results = evaluate(estimator=estimator, X=X, y=y, cv=cv, scoring=crps)\n",
    "\n",
    "# results are pd.DataFrame\n",
    "# each row is one repetition of the cross-validation on one fold fit/predict/evaluate\n",
    "# columns report performance, runtime, and other optional information (see docstring)\n",
    "results"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 3. Advanced composition patterns <a class=\"anchor\" id=\"chapter3\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "we introduce a number of composition patterns available in `skpro`:\n",
    "\n",
    "* reducer-wrappers that turn `sklearn` regressors into probabilistic ones\n",
    "* pipelines of `sklearn` transformers with `skpro` regressors\n",
    "* tuning `skpro` probabilistic regressors via grid/random search, minimizing a probabilistic metric\n",
    "* ensembling multiple `skpro` probabilistic regressors"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "data used in this section:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_diabetes\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X, y = load_diabetes(return_X_y=True, as_frame=True)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "evaluation metric used in this section:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "crps = CRPS()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1 Reducers to turn `sklearn` regressors probabilistic <a class=\"anchor\" id=\"section3_1\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "there are many common algorithms that turn a non-probabilistic tabular regressor probabilistic\n",
    "\n",
    "formally, this is a type of \"reduction\" - of probabilistic supervised tabular to non-probabilistic supervised tabular\n",
    "\n",
    "Examples:\n",
    "\n",
    "* predicting variance equal to training residual variance - `ResidualDouble` with standard settings\n",
    "    * or other unconditional distribution estimate for residuals\n",
    "* \"squaring the residual\" two-step prediction - `ResidualDouble`\n",
    "* bootstrap prediction intervals - `BootstrapRegressor`\n",
    "* conformal prediction intervals - `EnbpiRegressor`\n",
    "* MAPIE model agnostic prediction intervals - `MapieRegressor` (from `mapie` package)\n",
    "* natural gradient boosting aka NGBoost - `NGBoostRegressor` (from `ngboost` package)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This section presents some examples with visualizations.\n",
    "\n",
    "Not all of the above estimators are presented, but they can be used in the same way."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1.1 constant variance prediction <a class=\"anchor\" id=\"section3_1_1\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "30.508913426677797"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.model_selection import KFold\n",
    "\n",
    "# estimator specification - use any sklearn regressor for reg_mean\n",
    "reg_mean = RandomForestRegressor()\n",
    "reg_proba = ResidualDouble(reg_mean, cv=KFold(5))\n",
    "# cv is used to estimate out-of-sample residual variance via 5-fold CV\n",
    "# note - in-sample predictions will usually underestimate the variance!\n",
    "\n",
    "# fit and predict\n",
    "reg_proba.fit(X_train, y_train)\n",
    "y_pred_proba = reg_proba.predict_proba(X_test)\n",
    "\n",
    "# evaluate\n",
    "crps(y_test, y_pred_proba)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Correct label $y_i$', ylabel='Prediction interval $\\\\widehat{y}_i$'>"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from skpro.utils.plotting import plot_crossplot_interval\n",
    "\n",
    "plot_crossplot_interval(y_test, y_pred_proba, coverage=0.9)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1.2 two-step residual prediction <a class=\"anchor\" id=\"section3_1_2\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "30.22836313909981"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.model_selection import KFold\n",
    "\n",
    "# estimator specification - use any sklearn regressor for reg_mean and reg_resid\n",
    "reg_mean = RandomForestRegressor()\n",
    "reg_resid = RandomForestRegressor()\n",
    "reg_proba = ResidualDouble(reg_mean, estimator_resid=reg_resid, cv=KFold(5))\n",
    "# cv is used to estimate out-of-sample residual variance via 5-fold CV\n",
    "\n",
    "# fit and predict\n",
    "reg_proba.fit(X_train, y_train)\n",
    "y_pred_proba = reg_proba.predict_proba(X_test)\n",
    "\n",
    "# evaluate\n",
    "crps(y_test, y_pred_proba)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Correct label $y_i$', ylabel='Prediction interval $\\\\widehat{y}_i$'>"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from skpro.utils.plotting import plot_crossplot_interval\n",
    "\n",
    "plot_crossplot_interval(y_test, y_pred_proba, coverage=0.9)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.1.3 bootstrap prediction intervals <a class=\"anchor\" id=\"section3_1_3\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "35.5936968853179"
      ]
     },
     "execution_count": 64,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "from skpro.regression.bootstrap import BootstrapRegressor\n",
    "\n",
    "# estimator specification - use any sklearn regressor for reg_mean\n",
    "reg_mean = LinearRegression()\n",
    "reg_proba = BootstrapRegressor(reg_mean, n_bootstrap_samples=100)\n",
    "\n",
    "# fit and predict\n",
    "reg_proba.fit(X_train, y_train)\n",
    "y_pred_proba = reg_proba.predict_proba(X_test)\n",
    "\n",
    "# evaluate\n",
    "crps(y_test, y_pred_proba)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Axes: xlabel='Correct label $y_i$', ylabel='Prediction interval $\\\\widehat{y}_i$'>"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from skpro.utils.plotting import plot_crossplot_interval\n",
    "\n",
    "plot_crossplot_interval(y_test, y_pred_proba, coverage=0.9)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.2 Pipelines of `skpro` regressor and `sklearn` transformers <a class=\"anchor\" id=\"section3_2\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`skpro` regressors can be pipelined with `sklearn` transformers, using the `skpro` pipeline.\n",
    "\n",
    "This ensure presence of `predict_proba` etc in the pipeline object.\n",
    "\n",
    "The syntax is exactly the same as for `sklearn`'s pipeline."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_diabetes\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X, y = load_diabetes(return_X_y=True, as_frame=True)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.impute import SimpleImputer as Imputer\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "\n",
    "from skpro.regression.compose import Pipeline\n",
    "from skpro.regression.residual import ResidualDouble\n",
    "\n",
    "# estimator specification\n",
    "reg_mean = LinearRegression()\n",
    "reg_proba = ResidualDouble(reg_mean)\n",
    "\n",
    "# pipeline is specified as a list of tuples (name, estimator)\n",
    "pipe = Pipeline(\n",
    "    steps=[\n",
    "        (\"imputer\", Imputer()),  # an sklearn transformer\n",
    "        (\"scaler\", MinMaxScaler()),  # an sklearn transformer\n",
    "        (\"regressor\", reg_proba),  # an skpro regressor\n",
    "    ]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb {color: black;background-color: white;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb pre{padding: 0;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-toggleable {background-color: white;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-estimator:hover {background-color: #d4ebff;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-item {z-index: 1;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-parallel-item:only-child::after {width: 0;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb div.sk-text-repr-fallback {display: none;}</style><div id='sk-69394834-0275-4a85-b6cb-7a63a9c5c1cb' class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>Pipeline(steps=[(&#x27;imputer&#x27;, SimpleImputer()), (&#x27;scaler&#x27;, MinMaxScaler()),\n",
       "                (&#x27;regressor&#x27;, ResidualDouble(estimator=LinearRegression()))])</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class='sk-label-container'><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('2aa02a1d-986d-4ee7-b7ff-cd834e40de1b') type=\"checkbox\" ><label for=UUID('2aa02a1d-986d-4ee7-b7ff-cd834e40de1b') class='sk-toggleable__label sk-toggleable__label-arrow'>Pipeline</label><div class=\"sk-toggleable__content\"><pre>Pipeline(steps=[(&#x27;imputer&#x27;, SimpleImputer()), (&#x27;scaler&#x27;, MinMaxScaler()),\n",
       "                (&#x27;regressor&#x27;, ResidualDouble(estimator=LinearRegression()))])</pre></div></div></div><div class=\"sk-serial\"><div class='sk-item'><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('b0e0de17-4730-42a8-82d3-63790cb8e6c9') type=\"checkbox\" ><label for=UUID('b0e0de17-4730-42a8-82d3-63790cb8e6c9') class='sk-toggleable__label sk-toggleable__label-arrow'>SimpleImputer</label><div class=\"sk-toggleable__content\"><pre>SimpleImputer()</pre></div></div></div><div class='sk-item'><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('f5df8e0e-ae61-49c3-8ab6-76bfa7572292') type=\"checkbox\" ><label for=UUID('f5df8e0e-ae61-49c3-8ab6-76bfa7572292') class='sk-toggleable__label sk-toggleable__label-arrow'>MinMaxScaler</label><div class=\"sk-toggleable__content\"><pre>MinMaxScaler()</pre></div></div></div><div class=\"sk-item sk-dashed-wrapped\"><div class='sk-label-container'><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('f2b9fc99-99c0-4504-83f6-6a6db0437942') type=\"checkbox\" ><label for=UUID('f2b9fc99-99c0-4504-83f6-6a6db0437942') class='sk-toggleable__label sk-toggleable__label-arrow'>regressor: ResidualDouble</label><div class=\"sk-toggleable__content\"><pre>ResidualDouble(estimator=LinearRegression())</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class='sk-item'><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('1737408c-7d0c-48de-88c2-379439c0a8d2') type=\"checkbox\" ><label for=UUID('1737408c-7d0c-48de-88c2-379439c0a8d2') class='sk-toggleable__label sk-toggleable__label-arrow'>LinearRegression</label><div class=\"sk-toggleable__content\"><pre>LinearRegression()</pre></div></div></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "Pipeline(steps=[('imputer', SimpleImputer()), ('scaler', MinMaxScaler()),\n",
       "                ('regressor', ResidualDouble(estimator=LinearRegression()))])"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "# the pipeline behaves as any skpro regressor\n",
    "pipe.fit(X_train, y_train)\n",
    "y_pred = pipe.predict(X=X_test)\n",
    "y_pred_proba = pipe.predict_proba(X=X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "the pipeline provides the familiar nested `get_params`, `set_params` interface:\n",
    "\n",
    "nested parameters are keyed `componentname__parametername`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'steps': [('imputer', SimpleImputer()),\n",
       "  ('scaler', MinMaxScaler()),\n",
       "  ('regressor', ResidualDouble(estimator=LinearRegression()))],\n",
       " 'imputer': SimpleImputer(),\n",
       " 'scaler': MinMaxScaler(),\n",
       " 'regressor': ResidualDouble(estimator=LinearRegression()),\n",
       " 'imputer__add_indicator': False,\n",
       " 'imputer__copy': True,\n",
       " 'imputer__fill_value': None,\n",
       " 'imputer__keep_empty_features': False,\n",
       " 'imputer__missing_values': nan,\n",
       " 'imputer__strategy': 'mean',\n",
       " 'scaler__clip': False,\n",
       " 'scaler__copy': True,\n",
       " 'scaler__feature_range': (0, 1),\n",
       " 'regressor__cv': None,\n",
       " 'regressor__distr_loc_scale_name': None,\n",
       " 'regressor__distr_params': None,\n",
       " 'regressor__distr_type': 'Normal',\n",
       " 'regressor__estimator': LinearRegression(),\n",
       " 'regressor__estimator_resid': None,\n",
       " 'regressor__min_scale': 1e-10,\n",
       " 'regressor__residual_trafo': 'absolute',\n",
       " 'regressor__use_y_pred': False,\n",
       " 'regressor__estimator__copy_X': True,\n",
       " 'regressor__estimator__fit_intercept': True,\n",
       " 'regressor__estimator__n_jobs': None,\n",
       " 'regressor__estimator__positive': False}"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pipe.get_params()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "pipelines can also be created via simple lists of estimators,\n",
    "\n",
    "in this case names are generated automatically:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "# pipeline is specified as a list of tuples (name, estimator)\n",
    "pipe = Pipeline(\n",
    "    steps=[\n",
    "        Imputer(),  # an sklearn transformer\n",
    "        MinMaxScaler(),  # an sklearn transformer\n",
    "        reg_proba,  # an skpro regressor\n",
    "    ]\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.3 Tuning of `skpro` regressors via grid and random search <a class=\"anchor\" id=\"section3_3\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "`skpro` provides grid and random search tuners to tune arbitrary probabilistic regressors,\n",
    "\n",
    "using probabilistic metrics. Besides this, they function as the `sklearn` tuners do."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_diabetes\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X, y = load_diabetes(return_X_y=True, as_frame=True)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.model_selection import KFold\n",
    "\n",
    "from skpro.metrics import CRPS\n",
    "from skpro.model_selection import GridSearchCV\n",
    "from skpro.regression.residual import ResidualDouble\n",
    "\n",
    "# cross-validation specification for tuner\n",
    "cv = KFold(n_splits=3)\n",
    "\n",
    "# estimator to be tuned\n",
    "estimator = ResidualDouble(LinearRegression())\n",
    "\n",
    "# tuning grid - do we fit an intercept in the linear regression?\n",
    "param_grid = {\"estimator__fit_intercept\": [True, False]}\n",
    "\n",
    "# metric to be optimized\n",
    "crps_metric = CRPS()\n",
    "\n",
    "# specification of the grid search tuner\n",
    "gscv = GridSearchCV(\n",
    "    estimator=estimator,\n",
    "    param_grid=param_grid,\n",
    "    cv=cv,\n",
    "    scoring=crps_metric,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style>#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 {color: black;background-color: white;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 pre{padding: 0;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-toggleable {background-color: white;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-estimator:hover {background-color: #d4ebff;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-item {z-index: 1;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-parallel::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 2em;bottom: 0;left: 50%;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-parallel-item {display: flex;flex-direction: column;position: relative;background-color: white;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-parallel-item:only-child::after {width: 0;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;position: relative;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-label label {font-family: monospace;font-weight: bold;background-color: white;display: inline-block;line-height: 1.2em;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-label-container {position: relative;z-index: 2;text-align: center;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05 div.sk-text-repr-fallback {display: none;}</style><div id='sk-ded8fde5-f489-4fb3-8ad9-9dd06d30fa05' class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>GridSearchCV(cv=KFold(n_splits=3, random_state=None, shuffle=False),\n",
       "             estimator=ResidualDouble(estimator=LinearRegression()),\n",
       "             param_grid={&#x27;estimator__fit_intercept&#x27;: [True, False]},\n",
       "             scoring=CRPS())</pre><b>Please rerun this cell to show the HTML repr or trust the notebook.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item sk-dashed-wrapped\"><div class='sk-label-container'><div class=\"sk-label sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('be352d28-b0b7-4df9-a956-39e31f319727') type=\"checkbox\" ><label for=UUID('be352d28-b0b7-4df9-a956-39e31f319727') class='sk-toggleable__label sk-toggleable__label-arrow'>GridSearchCV</label><div class=\"sk-toggleable__content\"><pre>GridSearchCV(cv=KFold(n_splits=3, random_state=None, shuffle=False),\n",
       "             estimator=ResidualDouble(estimator=LinearRegression()),\n",
       "             param_grid={&#x27;estimator__fit_intercept&#x27;: [True, False]},\n",
       "             scoring=CRPS())</pre></div></div></div><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class=\"sk-item sk-dashed-wrapped\"><div class=\"sk-parallel\"><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class='sk-item'><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('b184a20b-70e5-4662-92a6-c5cb33e648f1') type=\"checkbox\" ><label for=UUID('b184a20b-70e5-4662-92a6-c5cb33e648f1') class='sk-toggleable__label sk-toggleable__label-arrow'>LinearRegression</label><div class=\"sk-toggleable__content\"><pre>LinearRegression()</pre></div></div></div></div></div></div></div></div></div></div></div><div class=\"sk-parallel-item\"><div class=\"sk-item\"><div class=\"sk-serial\"><div class='sk-item'><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=UUID('efed9304-f3f0-4f18-8d17-647e905c2c3e') type=\"checkbox\" ><label for=UUID('efed9304-f3f0-4f18-8d17-647e905c2c3e') class='sk-toggleable__label sk-toggleable__label-arrow'>CRPS</label><div class=\"sk-toggleable__content\"><pre>CRPS()</pre></div></div></div></div></div></div></div></div></div></div>"
      ],
      "text/plain": [
       "GridSearchCV(cv=KFold(n_splits=3, random_state=None, shuffle=False),\n",
       "             estimator=ResidualDouble(estimator=LinearRegression()),\n",
       "             param_grid={'estimator__fit_intercept': [True, False]},\n",
       "             scoring=CRPS())"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gscv"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "the grid search tuner behaves like any `skpro` probabilistic regressor:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "gscv.fit(X_train, y_train)\n",
    "y_pred = gscv.predict(X_test)\n",
    "y_pred_proba = gscv.predict_proba(X_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "random search is similar, except that instead of a grid a parameter sampler should be specified:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "from skpro.model_selection import RandomizedSearchCV\n",
    "\n",
    "# only difference to GridSearchCV is the param_distributions argument\n",
    "\n",
    "# specification of the random search parameter sampler\n",
    "param_distributions = {\"estimator__fit_intercept\": [True, False]}\n",
    "\n",
    "# specification of the random search tuner\n",
    "rscv = RandomizedSearchCV(\n",
    "    estimator=estimator,\n",
    "    param_distributions=param_distributions,\n",
    "    cv=cv,\n",
    "    scoring=crps_metric,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3.4 Bagging/mixture ensemble of probabilistic regressors <a class=\"anchor\" id=\"section3_3\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Classical bagging does the following, for a wrapped estimator:\n",
    "\n",
    "In `fit`:\n",
    "\n",
    "1. subsample rows and/or columns of `X`, `y` to `X_subs`, `y_subs`\n",
    "2. fit clone of wrapped estimator to `X_subs`, `y_subs`\n",
    "3. Repeat 1-2 `n_estimators` times, store that many fitted clones.\n",
    "\n",
    "In `predict`, for `X_test`:\n",
    "\n",
    "1. for all fitted clones, obtain predictions on `X_test` - these are distributions\n",
    "2. return the uniform mixture of these distributions, per test sample"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_diabetes\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X, y = load_diabetes(return_X_y=True, as_frame=True)\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "from skpro.regression.ensemble import BaggingRegressor\n",
    "from skpro.regression.residual import ResidualDouble\n",
    "\n",
    "reg_mean = LinearRegression()\n",
    "reg_proba = ResidualDouble(reg_mean)\n",
    "\n",
    "ens = BaggingRegressor(reg_proba, n_estimators=10)\n",
    "ens.fit(X_train, y_train)\n",
    "\n",
    "y_pred = ens.predict_proba(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "\"Mixture(columns=Index(['target'], dtype='object'),\\n        distributions=[Normal(columns=Index(['target'], dtype='object'),\\n                              index=Index([366, 311, 188, 263,  68,  96, 427, 235, 424, 236,\\n       ...\\n       132,  19, 173, 379, 227,  35, 401, 174,   5, 148],\\n      dtype='int64', length=111),\\n                              mu=array([[235.83179208],\\n       [163.58317773],\\n       [163.57602551],\\n       [ 39.19668912],\\n       [112.44346716],\\n       [214.38067207],\\n       [131.81838619],\\n       [186.17...\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649],\\n       [40.51030649]]))],\\n        index=Index([366, 311, 188, 263,  68,  96, 427, 235, 424, 236,\\n       ...\\n       132,  19, 173, 379, 227,  35, 401, 174,   5, 148],\\n      dtype='int64', length=111))\""
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# y_pred is a mixture distribution!\n",
    "str(y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[skpro.distributions.normal.Normal,\n",
       " skpro.distributions.normal.Normal,\n",
       " skpro.distributions.normal.Normal,\n",
       " skpro.distributions.normal.Normal,\n",
       " skpro.distributions.normal.Normal,\n",
       " skpro.distributions.normal.Normal,\n",
       " skpro.distributions.normal.Normal,\n",
       " skpro.distributions.normal.Normal,\n",
       " skpro.distributions.normal.Normal,\n",
       " skpro.distributions.normal.Normal]"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "[type(x) for x in y_pred.distributions]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 4. Extension guide - implementing your own probabilistic regressor <a class=\"anchor\" id=\"chapter4\"></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "`skpro` is meant to be easily extensible, for direct contribution to `skpro` as well as for local/private extension with custom methods.\n",
    "\n",
    "To get started:\n",
    "\n",
    "* Follow the [\"implementing estimator\" developer guide](https://skpro.readthedocs.io/en/stable/developer_guide/add_estimators.html)\n",
    "* Use the [probabilistic regressor template](https://github.com/sktime/skpro/blob/main/extension_templates/regression.py) to get started\n",
    "\n",
    "1. Read through the [probabilistic regression extension template](https://github.com/sktime/skpro/blob/main/extension_templates/regression.py) - this is a `python` file with `todo` blocks that mark the places in which changes need to be added.\n",
    "2. Copy the proba regressor extension template to a local folder in your own repository (local/private extension), or to a suitable location in your clone of the `skpro` or affiliated repository (if contributed extension), inside `skpro.regression`; rename the file and update the file docstring appropriately.\n",
    "3. Address the \"todo\" parts. Usually, this means: changing the name of the class, setting the tag values, specifying hyper-parameters, filling in `__init__`, `_fit`, and at least one of the probabilistic prediction methods, preferably `_predict_proba` (for details see the extension template). You can add private methods as long as they do not override the default public interface. For more details, see the extension template.\n",
    "4. To test your estimator manually: import your estimator and run it in the workflows in Section 1; then use it in the compositors in Section 3.\n",
    "5. To test your estimator automatically: call `skpro.utils.check_estimator` on your estimator. You can call this on a class or object instance. Ensure you have specified test parameters in the `get_test_params` method, according to the extension template.\n",
    "\n",
    "In case of direct contribution to `skpro` or one of its affiliated packages, additionally:\n",
    "\n",
    "* Add yourself as an author and/or a maintainer for the new estimator file(s), via `\"authors\"` and `\"maintainers\"` tag.\n",
    "* Create a pull request that contains only the new estimators (and their inheritance tree, if it's not just one class), as well as the automated tests as described above.\n",
    "* In the pull request, describe the estimator and optimally provide a publication or other technical reference for the strategy it implements.\n",
    "* Before making the pull request, ensure that you have all necessary permissions to contribute the code to a permissive license (BSD-3) open source project."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 5. Summary<a class=\"anchor\" id=\"chapter5\"></a>\n",
    "\n",
    "* `skpro` is a unified interface toolbox for probabilistic supervised regression, that is, for prediction intervals, quantiles, fully distributional predictions, in a tabular regression setting. The interface is fully interoperable with `scikit-learn` and `scikit-base` interface specifications.\n",
    "\n",
    "* `skpro` comes with rich composition functionality that allows to build complex pipelines easily, and connect easily with other parts of the open source ecosystem, such as `scikit-learn` and individual algorithm libraries.\n",
    "\n",
    "* `skpro` is easy to extend, and comes with user friendly tools to facilitate implementing and testing your own probabilistic regressors and composition principles."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "\n",
    "### Credits:\n",
    "\n",
    "noteook creation: fkiraly\n",
    "\n",
    "skpro: https://github.com/sktime/skpro/blob/main/CONTRIBUTORS.md"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "skpro",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.5"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "3e631b8a076cc106144e9b132b7d31cae2f1e2660b47e5f9fcb0397caae5fbd5"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}