Making n3fit use python pseudodata

n3fit/src/n3fit/scripts/n3fit_exec.py

@@ -26,10 +26,13 @@
     actions_ = []
 )
 
-N3FIT_PROVIDERS = ["n3fit.performfit",
-                   "validphys.results",
-                   "validphys.n3fit_data",
-                   "n3fit.n3fit_checks_provider"
+N3FIT_PROVIDERS = [
+    "n3fit.performfit",
+    "n3fit.n3fit_checks_provider",
+    "validphys.results",
+    "validphys.n3fit_data",
+    "validphys.pseudodata",
+    "validphys.commondata",
 ]
 
 log = logging.getLogger(__name__)

n3fit/src/n3fit/tests/regressions/quickcard_1.json

@@ -1,48 +1,97 @@
 {
-  "preprocessing": "",
+  "preprocessing": [
+    {
+      "fl": "sng",
+      "smallx": 1.1519856452941895,
+      "largex": 2.635816812515259,
+      "trainable": true
+    },
+    {
+      "fl": "g",
+      "smallx": 1.1843736171722412,
+      "largex": 1.5618072748184204,
+      "trainable": true
+    },
+    {
+      "fl": "v",
+      "smallx": 0.5399999022483826,
+      "largex": 2.0055015087127686,
+      "trainable": true
+    },
+    {
+      "fl": "v3",
+      "smallx": 0.3049614429473877,
+      "largex": 2.6231420040130615,
+      "trainable": true
+    },
+    {
+      "fl": "v8",
+      "smallx": 0.5784729719161987,
+      "largex": 2.1212551593780518,
+      "trainable": true
+    },
+    {
+      "fl": "t3",
+      "smallx": 1.3452119827270508,
+      "largex": 1.7576805353164673,
+      "trainable": true
+    },
+    {
+      "fl": "t8",
+      "smallx": 1.0689963102340698,
+      "largex": 1.9449496269226074,
+      "trainable": true
+    },
+    {
+      "fl": "cp",
+      "smallx": 0.741088330745697,
+      "largex": 3.4628548622131348,
+      "trainable": true
+    }
+  ],
   "stop_epoch": 1000,
   "best_epoch": 1000,
-  "erf_tr": 33.213114420572914,
-  "erf_vl": 28.218009317727912,
-  "chi2": 21.431137768190297,
+  "erf_tr": 35.36870388454861,
+  "erf_vl": 29.485028718647204,
+  "chi2": 21.601274632695894,
   "pos_state": "POS_VETO",
   "arc_lengths": [
-    1.1068653239049326,
-    1.4953119558272292,
-    1.201799272013087,
-    1.4270234891345712,
-    1.1920037459864075
+    1.1068650193928187,
+    1.495362525909499,
+    1.2016360471427414,
+    1.417299567086602,
+    1.1920902042035828
   ],
   "integrability": [
-    0.0026635357935483883,
-    0.0026635357935447246,
-    0.00014161983198945904,
-    5.059403657913208,
-    0.00397586848703213
+    0.002665607098606748,
+    0.002665607098605749,
+    0.00013836964808366936,
+    5.058645844459534,
+    0.0039739306957948806
   ],
   "timing": {
     "walltime": {
-      "Total": 14.067598581314087,
+      "Total": 41.31889986991882,
       "start": 0.0,
-      "data_loaded": 0.3445882797241211,
-      "replica_set": 0.34477806091308594,
-      "replica_fitted": 14.066914319992065,
-      "replica_set_to_replica_fitted": 13.72213625907898
+      "data_loaded": 0.033830881118774414,
+      "replica_set": 0.03406381607055664,
+      "replica_fitted": 41.31194734573364,
+      "replica_set_to_replica_fitted": 41.277883529663086
     },
     "cputime": {
-      "Total": 15.760098933,
+      "Total": 41.987736192999996,
       "start": 0.0,
-      "data_loaded": 1.3225217320000002,
-      "replica_set": 1.3239677300000001,
-      "replica_fitted": 15.759413147999998,
-      "replica_set_to_replica_fitted": 14.435445417999999
+      "data_loaded": 0.020405508000000516,
+      "replica_set": 0.020640056000000406,
+      "replica_fitted": 41.984591997,
+      "replica_set_to_replica_fitted": 41.963951941000005
     }
   },
   "version": {
-    "keras": "2.4.0",
-    "tensorflow": "2.3.1, mkl=False",
-    "numpy": "1.18.5",
-    "nnpdf": "3.5.1439+g47c1a11a6-dev",
-    "validphys": "3.5.1439+g47c1a11a6-dev"
+    "keras": "2.3.0-tf",
+    "tensorflow": "2.2.0, mkl=False",
+    "numpy": "1.19.1",
+    "nnpdf": "4.0.1.316+gf0c7b34",
+    "validphys": "4.0.1.316+gf0c7b34"
   }
-}
+}

n3fit/src/n3fit/tests/regressions/quickcard_2.json

@@ -1,48 +1,97 @@
 {
-  "preprocessing": "",
+  "preprocessing": [
+    {
+      "fl": "sng",
+      "smallx": 1.102711796760559,
+      "largex": 2.700000047683716,
+      "trainable": true
+    },
+    {
+      "fl": "g",
+      "smallx": 0.9399998784065247,
+      "largex": 1.6224322319030762,
+      "trainable": true
+    },
+    {
+      "fl": "v",
+      "smallx": 0.7402178645133972,
+      "largex": 1.7066106796264648,
+      "trainable": true
+    },
+    {
+      "fl": "v3",
+      "smallx": 0.21739234030246735,
+      "largex": 1.3615831136703491,
+      "trainable": true
+    },
+    {
+      "fl": "v8",
+      "smallx": 0.7599998712539673,
+      "largex": 2.411074161529541,
+      "trainable": true
+    },
+    {
+      "fl": "t3",
+      "smallx": 1.4385331869125366,
+      "largex": 2.29486083984375,
+      "trainable": true
+    },
+    {
+      "fl": "t8",
+      "smallx": 1.0592598915100098,
+      "largex": 1.7743843793869019,
+      "trainable": true
+    },
+    {
+      "fl": "cp",
+      "smallx": 0.22786737978458405,
+      "largex": 2.78507661819458,
+      "trainable": true
+    }
+  ],
   "stop_epoch": 1000,
-  "best_epoch": 998,
-  "erf_tr": 2.606570236771195,
-  "erf_vl": 3.0318090969458558,
-  "chi2": 1.5918478894589552,
+  "best_epoch": 999,
+  "erf_tr": 2.491444057888455,
+  "erf_vl": 2.8942466391656634,
+  "chi2": 1.6324648501268073,
   "pos_state": "POS_PASS",
   "arc_lengths": [
-    1.931689594079703,
-    1.697054671644156,
-    1.1340829034892161,
-    32.891455584502694,
-    1.1119670640240853
+    1.9364193786789174,
+    1.5682913617806982,
+    1.1399297208463033,
+    31.118964175987294,
+    1.1057977439528737
   ],
   "integrability": [
-    0.02683432819322168,
-    0.026834328193217516,
-    0.0004258475692022756,
-    20.328540325164795,
-    0.028318760450927627
+    0.02547394391149188,
+    0.025473943911492936,
+    0.00041527650682848805,
+    19.97365140914917,
+    0.02800018619745881
   ],
   "timing": {
     "walltime": {
-      "Total": 16.93633460998535,
+      "Total": 43.86756992340088,
       "start": 0.0,
-      "data_loaded": 0.34407711029052734,
-      "replica_set": 0.3442709445953369,
-      "replica_fitted": 16.935593366622925,
-      "replica_set_to_replica_fitted": 16.591322422027588
+      "data_loaded": 0.03478097915649414,
+      "replica_set": 0.035103559494018555,
+      "replica_fitted": 43.86035966873169,
+      "replica_set_to_replica_fitted": 43.82525610923767
     },
     "cputime": {
-      "Total": 18.655240567,
+      "Total": 44.583864212,
       "start": 0.0,
-      "data_loaded": 1.317259243,
-      "replica_set": 1.3187306250000002,
-      "replica_fitted": 18.65449747,
-      "replica_set_to_replica_fitted": 17.335766845000002
+      "data_loaded": 0.01997594100000022,
+      "replica_set": 0.020298879999999464,
+      "replica_fitted": 44.580285373,
+      "replica_set_to_replica_fitted": 44.559986493000004
     }
   },
   "version": {
-    "keras": "2.4.0",
-    "tensorflow": "2.3.1, mkl=False",
-    "numpy": "1.18.5",
-    "nnpdf": "3.5.1439+g47c1a11a6-dev",
-    "validphys": "3.5.1439+g47c1a11a6-dev"
+    "keras": "2.3.0-tf",
+    "tensorflow": "2.2.0, mkl=False",
+    "numpy": "1.19.1",
+    "nnpdf": "4.0.1.316+gf0c7b34-dev",
+    "validphys": "4.0.1.316+gf0c7b34-dev"
   }
-}
+}

validphys2/src/validphys/n3fit_data.py

@@ -205,7 +205,7 @@ def generate_data_replica(data, replica_mcseed):
 
 def fitting_data_dict(
     data,
-    generate_data_replica,
+    make_replica,
     tr_masks,
     kfold_masks,
     t0set=None,
@@ -258,7 +258,7 @@ def fitting_data_dict(
         t0pdfset = t0set.load_t0()
         spec_c.SetT0(t0pdfset)
 
-    expdata = generate_data_replica
+    expdata = make_replica
 
     datasets = common_data_reader_experiment(spec_c, data)
 

validphys2/src/validphys/pseudodata.py

@@ -115,7 +115,7 @@ def read_fit_pseudodata(fitcontext, context_index):
     return data_indices_list
 
 
-def make_replica(dataset_inputs_loaded_cd_with_cuts, seed=None):
+def make_replica(dataset_inputs_loaded_cd_with_cuts, replica_mcseed):
     """Function that takes in a list of :py:class:`validphys.coredata.CommonData`
     objects and returns a pseudodata replica accounting for
     possible correlations between systematic uncertainties.
@@ -153,7 +153,7 @@ def make_replica(dataset_inputs_loaded_cd_with_cuts, seed=None):
         0.34843355, 0.34730928, 0.3090914 , 0.32825111, 0.3485292 ,
     """
     # Seed the numpy RNG with the seed.
-    rng = np.random.default_rng(seed=seed)
+    rng = np.random.default_rng(seed=replica_mcseed)
 
     # The inner while True loop is for ensuring a positive definite
     # pseudodata replica

validphys2/src/validphys/tests/test_pythonmakereplica.py

@@ -16,6 +16,8 @@
 from validphys.tests.test_covmats import CORR_DATA
 
 
+SEED = 123456
+
 SINGLE_SYS_DATASETS = [
     {"dataset": "DYE886R"},
     {"dataset": "D0ZRAP", "cfac": ["QCD"]},
@@ -42,7 +44,7 @@ def test_commondata_unchanged(data_config, dataset_inputs, use_cuts):
     pre_mkrep_sys_tabs = [deepcopy(cd.systematics_table) for cd in ld_cds]
     pre_mkrep_cd_tabs = [deepcopy(cd.commondata_table) for cd in ld_cds]
 
-    make_replica(ld_cds)
+    make_replica(ld_cds, SEED)
 
     for post_mkrep_cd, pre_mkrep_cv in zip(ld_cds, pre_mkrep_cvs):
         assert_series_equal(post_mkrep_cd.central_values, pre_mkrep_cv)
@@ -63,13 +65,12 @@ def test_pseudodata_seeding(data_config, dataset_inputs, use_cuts):
     the random numbers are generated and used identically.
 
     """
-    seed = 123456
     config = dict(data_config)
     config["dataset_inputs"] = dataset_inputs
     config["use_cuts"] = use_cuts
-    ld_cds = API.dataset_inputs_loaded_cd_with_cuts(**config)
-    rep_1 = make_replica(ld_cds, seed=seed)
-    rep_2 = make_replica(ld_cds, seed=seed)
+    config["replica_mcseed"] = SEED
+    rep_1 = API.make_replica(**config)
+    rep_2 = API.make_replica(**config)
     np.testing.assert_allclose(rep_1, rep_2)
 
 
@@ -80,7 +81,8 @@ def test_pseudodata_has_correct_ndata(data_config, dataset_inputs, use_cuts):
     config = dict(data_config)
     config["dataset_inputs"] = dataset_inputs
     config["use_cuts"] = use_cuts
+    config["replica_mcseed"] = SEED
     ld_cds = API.dataset_inputs_loaded_cd_with_cuts(**config)
-    rep = make_replica(ld_cds)
+    rep = API.make_replica(**config)
     ndata = np.sum([cd.ndata for cd in ld_cds])
     assert len(rep) == ndata

validphys2/src/validphys/tests/test_regressions.py

@@ -61,7 +61,7 @@ def test_mcreplica(data_config):
     seed = 123456
     # Use no cuts because if filter rules change in the
     # future then this test will end up failing
-    rep = API.indexed_make_replica(**config, seed=seed)
+    rep = API.indexed_make_replica(**config, replica_mcseed=seed)
     return rep