roywei
require(pedquant)
require(tidytable)
require(purrr)
require(ggplot2)#获取数据整理数据
## 1/10 510050.ss
## 2/10 510300.ss
## 3/10 159919.sz
## 4/10 000300.ss
## 5/10 000852.ss
## 6/10 510500.ss
## 7/10 159915.sz
## 8/10 159922.sz
## 9/10 159901.sz
## 10/10 000016.ss
## Warning: `select.()` was deprecated in tidytable v0.10.0.
## ℹ Please use `select()` instead.
## ℹ Please note that all `verb.()` syntax has now been deprecated.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.
## Registered S3 method overwritten by 'quantmod':
## method from
## as.zoo.data.frame zoo
| Period | 510050.SS | 510300.SS | 159919.SZ | 000300.SS | 000852.SS | 510500.SS | 159915.SZ | 159922.SZ | 159901.SZ | 000016.SS |
|---|---|---|---|---|---|---|---|---|---|---|
| 2020/04/07 0:00:00 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| 2020/04/08 0:00:00 | -0.004751 | -0.003976 | -0.003912 | -0.004666 | 0.004023 | 0.002985 | -0.002114 | 0.002693 | -0.00493 | -0.004915 |
| 2020/04/09 0:00:00 | 0.003291 | 0.002917 | 0.004432 | 0.003293 | 0.01236 | 0.00925 | 0.01627 | 0.009812 | 0.003355 | 0.003166 |
| 2020/04/10 0:00:00 | -0.0003652 | -0.003183 | -0.004955 | -0.00625 | -0.02656 | -0.01823 | -0.02317 | -0.019 | -0.005533 | -0.0009959 |
| 2020/04/13 0:00:00 | -0.004392 | -0.005595 | -0.005505 | -0.004233 | -0.01046 | -0.01031 | -0.01287 | -0.009818 | -0.004568 | -0.003891 |
| 2020/04/14 0:00:00 | 0.01637 | 0.01643 | 0.01643 | 0.01912 | 0.02008 | 0.02122 | 0.02977 | 0.02079 | 0.01893 | 0.01685 |
数据整理宽数据变长数据，并绘制图 ## 数据描述统计
ret1 %>%
psych::describe()## vars n mean sd median trimmed mad min max range skew
## 510050.SS 1 728 0 0.01 0 0 0.01 -0.05 0.08 0.14 0.29
## 510300.SS 2 728 0 0.01 0 0 0.01 -0.05 0.07 0.12 -0.01
## 159919.SZ 3 728 0 0.01 0 0 0.01 -0.05 0.07 0.12 0.07
## 000300.SS 4 728 0 0.01 0 0 0.01 -0.05 0.06 0.11 -0.19
## 000852.SS 5 728 0 0.01 0 0 0.01 -0.08 0.05 0.13 -0.60
## 510500.SS 6 728 0 0.01 0 0 0.01 -0.14 0.05 0.18 -2.13
## 159915.SZ 7 727 0 0.02 0 0 0.02 -0.06 0.05 0.12 -0.12
## 159922.SZ 8 728 0 0.01 0 0 0.01 -0.10 0.05 0.15 -1.32
## 159901.SZ 9 727 0 0.03 0 0 0.01 -0.71 0.05 0.76 -18.10
## 000016.SS 10 728 0 0.01 0 0 0.01 -0.05 0.07 0.12 0.03
## kurtosis se
## 510050.SS 4.19 0
## 510300.SS 2.93 0
## 159919.SZ 3.19 0
## 000300.SS 1.89 0
## 000852.SS 2.32 0
## 510500.SS 19.20 0
## 159915.SZ 0.70 0
## 159922.SZ 8.89 0
## 159901.SZ 424.21 0
## 000016.SS 2.58 0
ret1 %>%
summarytools::descr()## Descriptive Statistics
##
## 000016.SS 000300.SS 000852.SS 159901.SZ 159915.SZ 159919.SZ
## ----------------- ----------- ----------- ----------- ----------- ----------- -----------
## Mean 0.00 0.00 0.00 0.00 0.00 0.00
## Std.Dev 0.01 0.01 0.01 0.03 0.02 0.01
## Min -0.05 -0.05 -0.08 -0.71 -0.06 -0.05
## Q1 -0.01 -0.01 -0.01 -0.01 -0.01 -0.01
## Median 0.00 0.00 0.00 0.00 0.00 0.00
## Q3 0.01 0.01 0.01 0.01 0.01 0.01
## Max 0.07 0.06 0.05 0.05 0.05 0.07
## MAD 0.01 0.01 0.01 0.01 0.02 0.01
## IQR 0.01 0.01 0.02 0.02 0.02 0.01
## CV -297.96 113.29 50.75 -43.94 57.71 106.47
## Skewness 0.03 -0.19 -0.60 -18.10 -0.12 0.07
## SE.Skewness 0.09 0.09 0.09 0.09 0.09 0.09
## Kurtosis 2.58 1.89 2.32 424.21 0.70 3.19
## N.Valid 728.00 728.00 728.00 727.00 727.00 728.00
## Pct.Valid 100.00 100.00 100.00 99.86 99.86 100.00
##
## Table: Table continues below
##
##
##
## 159922.SZ 510050.SS 510300.SS 510500.SS
## ----------------- ----------- ----------- ----------- -----------
## Mean 0.00 0.00 0.00 0.00
## Std.Dev 0.01 0.01 0.01 0.01
## Min -0.10 -0.05 -0.05 -0.14
## Q1 -0.01 -0.01 -0.01 -0.01
## Median 0.00 0.00 0.00 0.00
## Q3 0.01 0.01 0.01 0.01
## Max 0.05 0.08 0.07 0.05
## MAD 0.01 0.01 0.01 0.01
## IQR 0.01 0.01 0.01 0.01
## CV 54.55 -401.91 108.16 76.25
## Skewness -1.32 0.29 -0.01 -2.13
## SE.Skewness 0.09 0.09 0.09 0.09
## Kurtosis 8.89 4.19 2.93 19.20
## N.Valid 728.00 728.00 728.00 728.00
## Pct.Valid 100.00 100.00 100.00 100.00
ret1 %>%
summarytools::dfSummary()## . was converted to a data frame
## Data Frame Summary
## .
## Dimensions: 728 x 10
## Duplicates: 0
##
## ---------------------------------------------------------------------------------------------------------
## No Variable Stats / Values Freqs (% of Valid) Graph Valid Missing
## ---- ----------- ----------------------- --------------------- --------------------- ---------- ---------
## 1 510050.SS Mean (sd) : 0 (0) 716 distinct values : 728 0
## [numeric] min < med < max: : . (100.0%) (0.0%)
## -0.1 < 0 < 0.1 : :
## IQR (CV) : 0 (-401.9) : :
## : : : .
##
## 2 510300.SS Mean (sd) : 0 (0) 722 distinct values . : 728 0
## [numeric] min < med < max: : : (100.0%) (0.0%)
## 0 < 0 < 0.1 : :
## IQR (CV) : 0 (108.2) : : .
## : : : : .
##
## 3 159919.SZ Mean (sd) : 0 (0) 716 distinct values . : 728 0
## [numeric] min < med < max: : : (100.0%) (0.0%)
## 0 < 0 < 0.1 : :
## IQR (CV) : 0 (106.5) : :
## : : : :
##
## 4 000300.SS Mean (sd) : 0 (0) 728 distinct values : 728 0
## [numeric] min < med < max: : . (100.0%) (0.0%)
## -0.1 < 0 < 0.1 : :
## IQR (CV) : 0 (113.3) : : :
## . : : : : .
##
## 5 000852.SS Mean (sd) : 0 (0) 728 distinct values : 728 0
## [numeric] min < med < max: : (100.0%) (0.0%)
## -0.1 < 0 < 0.1 : :
## IQR (CV) : 0 (50.8) : : :
## . : : : : .
##
## 6 510500.SS Mean (sd) : 0 (0) 723 distinct values : 728 0
## [numeric] min < med < max: : (100.0%) (0.0%)
## -0.1 < 0 < 0 :
## IQR (CV) : 0 (76.2) :
## : : :
##
## 7 159915.SZ Mean (sd) : 0 (0) 720 distinct values : 727 1
## [numeric] min < med < max: : : . (99.9%) (0.1%)
## -0.1 < 0 < 0.1 : : :
## IQR (CV) : 0 (57.7) . : : : :
## . : : : : : .
##
## 8 159922.SZ Mean (sd) : 0 (0) 719 distinct values . : 728 0
## [numeric] min < med < max: : : (100.0%) (0.0%)
## -0.1 < 0 < 0 : :
## IQR (CV) : 0 (54.6) : :
## : : : :
##
## 9 159901.SZ Mean (sd) : 0 (0) 720 distinct values : 727 1
## [numeric] min < med < max: : (99.9%) (0.1%)
## -0.7 < 0 < 0 :
## IQR (CV) : 0 (-43.9) :
## :
##
## 10 000016.SS Mean (sd) : 0 (0) 728 distinct values : 728 0
## [numeric] min < med < max: : (100.0%) (0.0%)
## -0.1 < 0 < 0.1 . : .
## IQR (CV) : 0 (-298) : : :
## . : : : .
## ---------------------------------------------------------------------------------------------------------
ret1 %>%
skimr::skim()| Name | Piped data |
| Number of rows | 728 |
| Number of columns | 10 |
| _______________________ | |
| Column type frequency: | |
| numeric | 10 |
| ________________________ | |
| Group variables | None |
Data summary
Variable type: numeric
| skim_variable | n_missing | complete_rate | mean | sd | p0 | p25 | p50 | p75 | p100 | hist |
|---|---|---|---|---|---|---|---|---|---|---|
| 510050.SS | 0 | 1 | 0 | 0.01 | -0.05 | -0.01 | 0 | 0.01 | 0.08 | ▁▇▆▁▁ |
| 510300.SS | 0 | 1 | 0 | 0.01 | -0.05 | -0.01 | 0 | 0.01 | 0.07 | ▁▆▇▁▁ |
| 159919.SZ | 0 | 1 | 0 | 0.01 | -0.05 | -0.01 | 0 | 0.01 | 0.07 | ▁▇▇▁▁ |
| 000300.SS | 0 | 1 | 0 | 0.01 | -0.05 | -0.01 | 0 | 0.01 | 0.06 | ▁▂▇▂▁ |
| 000852.SS | 0 | 1 | 0 | 0.01 | -0.08 | -0.01 | 0 | 0.01 | 0.05 | ▁▁▅▇▁ |
| 510500.SS | 0 | 1 | 0 | 0.01 | -0.14 | -0.01 | 0 | 0.01 | 0.05 | ▁▁▁▇▂ |
| 159915.SZ | 1 | 1 | 0 | 0.02 | -0.06 | -0.01 | 0 | 0.01 | 0.05 | ▁▂▇▅▁ |
| 159922.SZ | 0 | 1 | 0 | 0.01 | -0.10 | -0.01 | 0 | 0.01 | 0.05 | ▁▁▁▇▁ |
| 159901.SZ | 1 | 1 | 0 | 0.03 | -0.71 | -0.01 | 0 | 0.01 | 0.05 | ▁▁▁▁▇ |
| 000016.SS | 0 | 1 | 0 | 0.01 | -0.05 | -0.01 | 0 | 0.01 | 0.07 | ▁▃▇▁▁ |
ret1 %>%
pastecs::stat.desc(norm = T)## 510050.SS 510300.SS 159919.SZ 000300.SS
## nbr.val 7.280000e+02 7.280000e+02 7.280000e+02 7.280000e+02
## nbr.null 1.000000e+01 7.000000e+00 1.000000e+01 1.000000e+00
## nbr.na 0.000000e+00 0.000000e+00 0.000000e+00 0.000000e+00
## min -5.255785e-02 -4.937696e-02 -4.978692e-02 -5.068431e-02
## max 8.478633e-02 7.080112e-02 7.448666e-02 5.512915e-02
## range 1.373442e-01 1.201781e-01 1.242736e-01 1.058135e-01
## sum -2.248943e-02 8.126296e-02 8.192297e-02 7.726534e-02
## median 0.000000e+00 0.000000e+00 2.041037e-04 2.348117e-04
## mean -3.089207e-05 1.116250e-04 1.125316e-04 1.061337e-04
## SE.mean 4.601628e-04 4.474809e-04 4.440463e-04 4.456344e-04
## CI.mean.0.95 9.034066e-04 8.785090e-04 8.717662e-04 8.748839e-04
## var 1.541539e-04 1.457741e-04 1.435450e-04 1.445735e-04
## std.dev 1.241587e-02 1.207369e-02 1.198103e-02 1.202387e-02
## coef.var -4.019113e+02 1.081630e+02 1.064681e+02 1.132899e+02
## skewness 2.947560e-01 -7.364071e-03 6.907461e-02 -1.940789e-01
## skew.2SE 1.626728e+00 -4.064155e-02 3.812156e-01 -1.071101e+00
## kurtosis 4.189660e+00 2.929364e+00 3.188576e+00 1.886044e+00
## kurt.2SE 1.157687e+01 8.094418e+00 8.810673e+00 5.211515e+00
## normtest.W 9.623622e-01 9.699960e-01 9.697291e-01 9.803345e-01
## normtest.p 1.002531e-12 4.580695e-11 3.967789e-11 2.550417e-08
## 000852.SS 510500.SS 159915.SZ 159922.SZ
## nbr.val 7.280000e+02 7.280000e+02 7.270000e+02 7.280000e+02
## nbr.null 1.000000e+00 5.000000e+00 5.000000e+00 9.000000e+00
## nbr.na 0.000000e+00 0.000000e+00 1.000000e+00 0.000000e+00
## min -8.132766e-02 -1.363053e-01 -6.448925e-02 -1.023826e-01
## max 5.074896e-02 4.644455e-02 5.352141e-02 4.506149e-02
## range 1.320766e-01 1.827499e-01 1.180107e-01 1.474441e-01
## sum 1.996065e-01 1.225009e-01 2.157243e-01 1.630221e-01
## median 1.145565e-03 8.632768e-04 -3.669052e-04 9.948805e-04
## mean 2.741848e-04 1.682704e-04 2.967322e-04 2.239315e-04
## SE.mean 5.157277e-04 4.755150e-04 6.350645e-04 4.527600e-04
## CI.mean.0.95 1.012493e-03 9.335465e-04 1.246782e-03 8.888730e-04
## var 1.936298e-04 1.646114e-04 2.932041e-04 1.492339e-04
## std.dev 1.391509e-02 1.283010e-02 1.712320e-02 1.221613e-02
## coef.var 5.075078e+01 7.624690e+01 5.770591e+01 5.455299e+01
## skewness -6.026873e-01 -2.128324e+00 -1.174837e-01 -1.316797e+00
## skew.2SE -3.326169e+00 -1.174600e+01 -6.479368e-01 -7.267267e+00
## kurtosis 2.324962e+00 1.920499e+01 7.004449e-01 8.887326e+00
## kurt.2SE 6.424334e+00 5.306722e+01 1.934148e+00 2.455746e+01
## normtest.W 9.736253e-01 8.819846e-01 9.933844e-01 9.228234e-01
## normtest.p 3.521709e-10 4.311004e-23 2.699818e-03 7.545511e-19
## 159901.SZ 000016.SS
## nbr.val 7.270000e+02 7.280000e+02
## nbr.null 7.000000e+00 1.000000e+00
## nbr.na 1.000000e+00 0.000000e+00
## min -7.140083e-01 -5.367165e-02
## max 4.599049e-02 6.583431e-02
## range 7.599987e-01 1.195060e-01
## sum -4.999985e-01 -2.999859e-02
## median -2.493455e-04 1.743727e-05
## mean -6.877558e-04 -4.120685e-05
## SE.mean 1.120815e-03 4.550520e-04
## CI.mean.0.95 2.200426e-03 8.933729e-04
## var 9.132766e-04 1.507487e-04
## std.dev 3.022047e-02 1.227797e-02
## coef.var -4.394069e+01 -2.979596e+02
## skewness -1.809841e+01 2.709624e-02
## skew.2SE -9.981491e+01 1.495413e-01
## kurtosis 4.242081e+02 2.581013e+00
## kurt.2SE 1.171372e+03 7.131855e+00
## normtest.W 3.387740e-01 9.736936e-01
## normtest.p 4.801377e-45 3.665486e-10
ret1 %>%timetk::tk_tbl(rename_index = 'Date') %>%
tidytable::pivot_longer(cols = -Date,names_to = 'stock',values_to = 'return')->ret_long
ggplot2::ggplot(ret_long,mapping = aes(x=Date,y=return,color=stock))+geom_path(stat = "identity") +
facet_grid(stock ~ .) + theme_minimal() + labs(x = "Date", y = "returns")
This is an R Markdown format used for publishing markdown documents to GitHub. When you click the Knit button all R code chunks are run and a markdown file (.md) suitable for publishing to GitHub is generated.
You can include R code in the document as follows:
summary(cars)## speed dist
## Min. : 4.0 Min. : 2.00
## 1st Qu.:12.0 1st Qu.: 26.00
## Median :15.0 Median : 36.00
## Mean :15.4 Mean : 42.98
## 3rd Qu.:19.0 3rd Qu.: 56.00
## Max. :25.0 Max. :120.00
You can also embed plots, for example:
Note that the echo = FALSE parameter was added to the code chunk to
prevent printing of the R code that generated the plot.
数据清理，生成变量以便于机器学习使用
require(tidytable)
require(tidyquant)## 载入需要的程辑包：tidyquant
## 载入需要的程辑包：lubridate
##
## 载入程辑包：'lubridate'
## The following objects are masked from 'package:base':
##
## date, intersect, setdiff, union
## 载入需要的程辑包：PerformanceAnalytics
## 载入需要的程辑包：xts
## 载入需要的程辑包：zoo
##
## 载入程辑包：'zoo'
## The following objects are masked from 'package:base':
##
## as.Date, as.Date.numeric
##
## ################################### WARNING ###################################
## # We noticed you have dplyr installed. The dplyr lag() function breaks how #
## # base R's lag() function is supposed to work, which breaks lag(my_xts). #
## # #
## # If you call library(dplyr) later in this session, then calls to lag(my_xts) #
## # that you enter or source() into this session won't work correctly. #
## # #
## # All package code is unaffected because it is protected by the R namespace #
## # mechanism. #
## # #
## # Set `options(xts.warn_dplyr_breaks_lag = FALSE)` to suppress this warning. #
## # #
## # You can use stats::lag() to make sure you're not using dplyr::lag(), or you #
## # can add conflictRules('dplyr', exclude = 'lag') to your .Rprofile to stop #
## # dplyr from breaking base R's lag() function. #
## ################################### WARNING ###################################
##
## 载入程辑包：'xts'
## The following objects are masked from 'package:tidytable':
##
## first, last
##
## 载入程辑包：'PerformanceAnalytics'
## The following object is masked from 'package:graphics':
##
## legend
## 载入需要的程辑包：quantmod
## 载入需要的程辑包：TTR
require(pedquant)
test <- md_stock("510500.ss", date_range = "10y") %>% bind_rows()## 1/1 510500.ss
test %>%
tq_mutate(
select = close_adj,
mutate_fun = SMA,
n = 5,
col_rename = "sma5"
) %>%
tq_mutate(
select = close_adj,
mutate_fun = EMA,
n = 5,
col_rename = "ema5"
) %>%
tq_mutate(
select = close_adj,
mutate_fun = RSI,
n = 5,
col_rename = "rsi"
) %>%
tq_mutate(
select = close_adj,
mutate_fun = dailyReturn,
type = "log",
col_rename = "ret1"
) -> test1
test %>%
tq_mutate(
select = close_adj,
mutate_fun = SMA,
n = 5,
col_rename = "sma5"
) %>%
tq_mutate(
select = close_adj,
mutate_fun = EMA,
n = 5,
col_rename = "ema5"
) %>%
tq_mutate(
select = close_adj,
mutate_fun = RSI,
n = 5,
col_rename = "rsi"
) %>%
tq_mutate(
select = close_adj,
mutate_fun = dailyReturn,
type = "log",
col_rename = "ret1"
) -> test1
test1 %>%
tq_mutate(
select = close_adj,
mutate_fun = MACD,
) %>%
mutate(Direction = ifelse(close_adj >= lag(close_adj, 5), 1, 0)) %>%
select(9:ncol(.)) %>%
mutate(sma5 = lag(sma5), ema5 = lag(ema5), rsi = lag(rsi), ret1 = lag(ret1), sigal = lag(signal), macd = lag(macd)) %>%
select(Direction, ret1, sma5, ema5, macd, signal, rsi) %>%
drop_na() %>%
`colnames<-`(c(
"Direction", "Ret", "SMA", "EMA", "MACD", "Signal",
"RSI"
)) -> d_ex1
# 使用逻辑斯蒂回归进行分类预测
# remove NAs
d_ex1 <- na.omit(d_ex1)
# convert to data frame
d_ex1 <- as.data.frame(d_ex1)
# convert direction to a factor for classification
d_ex1$Direction <- as.factor(d_ex1$Direction)
idx1 <- c(1:round(nrow(d_ex1) * 0.7)) # create index for first 70% values to be in the testing set
d_train1 <- d_ex1[idx1, ] # training set
d_test1 <- d_ex1[-idx1, ] # testing set
library(caret)## Warning: 程辑包'caret'是用R版本4.2.3 来建造的
## 载入需要的程辑包：lattice
##
## 载入程辑包：'caret'
## The following object is masked from 'package:purrr':
##
## lift
set.seed(999)
# control
cntrl1 <- trainControl(
method = "timeslice", initialWindow = 250, horizon = 30,
fixedWindow = TRUE
)
# preprocesing
prep1 <- c("center", "scale")
# logistic regression
logit_ex1 <- train(Direction ~ .,
data = d_train1, method = "glm", family = "binomial",
trControl = cntrl1, preProcess = prep1
)
logit_ex1 # final model accuracy## Generalized Linear Model
##
## 1678 samples
## 6 predictor
## 2 classes: '0', '1'
##
## Pre-processing: centered (6), scaled (6)
## Resampling: Rolling Forecasting Origin Resampling (30 held-out with a fixed window)
## Summary of sample sizes: 250, 250, 250, 250, 250, 250, ...
## Resampling results:
##
## Accuracy Kappa
## 0.7983798 0.5313487
summary(logit_ex1$finalModel) # summary of the final model##
## Call:
## NULL
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -3.3464 -0.5509 0.2589 0.6135 2.6190
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.45055 0.06864 6.564 5.23e-11 ***
## Ret -0.15775 0.09132 -1.727 0.0841 .
## SMA -33.08997 4.76512 -6.944 3.81e-12 ***
## EMA 32.98627 4.76204 6.927 4.30e-12 ***
## MACD -0.31932 0.49670 -0.643 0.5203
## Signal 0.19310 0.44862 0.430 0.6669
## RSI 1.97588 0.14794 13.356 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 2288.8 on 1677 degrees of freedom
## Residual deviance: 1367.2 on 1671 degrees of freedom
## AIC: 1381.2
##
## Number of Fisher Scoring iterations: 5
# 预测变量重要图示
library(vip)##
## 载入程辑包：'vip'
## The following object is masked from 'package:utils':
##
## vi
vip(logit_ex1, geom = "point") + theme_minimal()
# 使用混淆矩阵分析预测精度
pred1 <- predict(logit_ex1, newdata = d_test1) # prediction on the test data
confusionMatrix(data = pred1, reference = d_test1$Direction)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 245 45
## 1 78 351
##
## Accuracy : 0.8289
## 95% CI : (0.7994, 0.8558)
## No Information Rate : 0.5508
## P-Value [Acc > NIR] : < 2e-16
##
## Kappa : 0.651
##
## Mcnemar's Test P-Value : 0.00391
##
## Sensitivity : 0.7585
## Specificity : 0.8864
## Pos Pred Value : 0.8448
## Neg Pred Value : 0.8182
## Prevalence : 0.4492
## Detection Rate : 0.3408
## Detection Prevalence : 0.4033
## Balanced Accuracy : 0.8224
##
## 'Positive' Class : 0
##