diff --git a/Data/APLpredictor/AdaBoostRegressor/model.pkl b/Data/APLpredictor/AdaBoostRegressor/model.pkl
index 1804999d..be7802fa 100644
Binary files a/Data/APLpredictor/AdaBoostRegressor/model.pkl and b/Data/APLpredictor/AdaBoostRegressor/model.pkl differ
diff --git a/Data/APLpredictor/DecisionTreeRegressor/model.pkl b/Data/APLpredictor/DecisionTreeRegressor/model.pkl
index 0de558a6..afe5c28f 100644
Binary files a/Data/APLpredictor/DecisionTreeRegressor/model.pkl and b/Data/APLpredictor/DecisionTreeRegressor/model.pkl differ
diff --git a/Data/APLpredictor/ElasticNet/model.pkl b/Data/APLpredictor/ElasticNet/model.pkl
index 1f868d54..41f280a0 100644
Binary files a/Data/APLpredictor/ElasticNet/model.pkl and b/Data/APLpredictor/ElasticNet/model.pkl differ
diff --git a/Data/APLpredictor/GradientBoostingRegressor/model.pkl b/Data/APLpredictor/GradientBoostingRegressor/model.pkl
index dca861a0..8cbd3d71 100644
Binary files a/Data/APLpredictor/GradientBoostingRegressor/model.pkl and b/Data/APLpredictor/GradientBoostingRegressor/model.pkl differ
diff --git a/Data/APLpredictor/KNeighborsRegressor/model.pkl b/Data/APLpredictor/KNeighborsRegressor/model.pkl
index 009c00bb..b0220722 100644
Binary files a/Data/APLpredictor/KNeighborsRegressor/model.pkl and b/Data/APLpredictor/KNeighborsRegressor/model.pkl differ
diff --git a/Data/APLpredictor/Lasso/model.pkl b/Data/APLpredictor/Lasso/model.pkl
index 4a1ce68d..e82a5d57 100644
Binary files a/Data/APLpredictor/Lasso/model.pkl and b/Data/APLpredictor/Lasso/model.pkl differ
diff --git a/Data/APLpredictor/LinearRegression/model.pkl b/Data/APLpredictor/LinearRegression/model.pkl
index f2812fdd..f98b1154 100644
Binary files a/Data/APLpredictor/LinearRegression/model.pkl and b/Data/APLpredictor/LinearRegression/model.pkl differ
diff --git a/Data/APLpredictor/MLPRegressor/model.pkl b/Data/APLpredictor/MLPRegressor/model.pkl
index 5fd09c04..d8a0bb33 100644
Binary files a/Data/APLpredictor/MLPRegressor/model.pkl and b/Data/APLpredictor/MLPRegressor/model.pkl differ
diff --git a/Data/APLpredictor/RandomForestRegressor/model.pkl b/Data/APLpredictor/RandomForestRegressor/model.pkl
index 54f37a79..35662c36 100644
Binary files a/Data/APLpredictor/RandomForestRegressor/model.pkl and b/Data/APLpredictor/RandomForestRegressor/model.pkl differ
diff --git a/Data/APLpredictor/Ridge/model.pkl b/Data/APLpredictor/Ridge/model.pkl
index 2dbe22a1..11579448 100644
Binary files a/Data/APLpredictor/Ridge/model.pkl and b/Data/APLpredictor/Ridge/model.pkl differ
diff --git a/Data/APLpredictor/XGBRegressor/model.pkl b/Data/APLpredictor/XGBRegressor/model.pkl
index 7686743b..76865ebe 100644
Binary files a/Data/APLpredictor/XGBRegressor/model.pkl and b/Data/APLpredictor/XGBRegressor/model.pkl differ
diff --git a/Data/APLpredictor/gridsearch_log.txt b/Data/APLpredictor/gridsearch_log.txt
index 3106e43e..dbaf96ec 100644
--- a/Data/APLpredictor/gridsearch_log.txt
+++ b/Data/APLpredictor/gridsearch_log.txt
@@ -2116,3 +2116,938 @@ RMSE: 4.633
R2: 0.723
Pearson: 0.862
+---- LinearRegression ----
+
+Best model parameters: {}
+GridSearchCV fit: -4.673054513199252
+
+Metrics in test set
+MAE: 3.395
+RMSE: 4.673
+R2: 0.644
+Pearson: 0.802
+
+Metrics in training set
+MAE: 2.940
+RMSE: 3.743
+R2: 0.819
+Pearson: 0.905
+
+---- Lasso ----
+
+Best model parameters: {'alpha': 0.001}
+GridSearchCV fit: -4.673851754917777
+
+Metrics in test set
+MAE: 3.397
+RMSE: 4.674
+R2: 0.644
+Pearson: 0.802
+
+Metrics in training set
+MAE: 2.948
+RMSE: 3.745
+R2: 0.819
+Pearson: 0.905
+
+---- Ridge ----
+
+Best model parameters: {'alpha': 0.01}
+GridSearchCV fit: -4.6732333376506965
+
+Metrics in test set
+MAE: 3.396
+RMSE: 4.673
+R2: 0.644
+Pearson: 0.802
+
+Metrics in training set
+MAE: 2.948
+RMSE: 3.745
+R2: 0.819
+Pearson: 0.905
+
+---- ElasticNet ----
+
+Best model parameters: {'alpha': 0.05, 'l1_ratio': 0.9}
+GridSearchCV fit: -5.36241111391284
+
+Metrics in test set
+MAE: 4.036
+RMSE: 5.362
+R2: 0.531
+Pearson: 0.758
+
+Metrics in training set
+MAE: 3.825
+RMSE: 6.789
+R2: 0.404
+Pearson: 0.669
+
+---- DecisionTreeRegressor ----
+
+Best model parameters: {'max_depth': 10}
+GridSearchCV fit: -3.7898251391858016
+
+Metrics in test set
+MAE: 2.601
+RMSE: 3.790
+R2: 0.766
+Pearson: 0.875
+
+Metrics in training set
+MAE: 3.009
+RMSE: 4.739
+R2: 0.710
+Pearson: 0.849
+
+---- RandomForestRegressor ----
+
+Best model parameters: {'max_depth': None, 'max_features': 1, 'min_samples_leaf': 1, 'min_samples_split': 2, 'n_estimators': 100}
+GridSearchCV fit: -3.815245350355903
+
+Metrics in test set
+MAE: 2.655
+RMSE: 3.815
+R2: 0.762
+Pearson: 0.874
+
+Metrics in training set
+MAE: 2.853
+RMSE: 4.966
+R2: 0.681
+Pearson: 0.839
+
+---- GradientBoostingRegressor ----
+
+Best model parameters: {'learning_rate': 0.1, 'max_depth': 5, 'n_estimators': 75}
+GridSearchCV fit: -3.759779220340636
+
+Metrics in test set
+MAE: 2.593
+RMSE: 3.760
+R2: 0.769
+Pearson: 0.877
+
+Metrics in training set
+MAE: 2.785
+RMSE: 4.421
+R2: 0.747
+Pearson: 0.876
+
+---- AdaBoostRegressor ----
+
+Best model parameters: {'learning_rate': 0.01, 'n_estimators': 100}
+GridSearchCV fit: -4.501795574406243
+
+Metrics in test set
+MAE: 3.312
+RMSE: 4.502
+R2: 0.669
+Pearson: 0.818
+
+Metrics in training set
+MAE: 3.109
+RMSE: 4.731
+R2: 0.711
+Pearson: 0.852
+
+---- KNeighborsRegressor ----
+
+Best model parameters: {'n_neighbors': 10}
+GridSearchCV fit: -4.640041949298464
+
+Metrics in test set
+MAE: 3.142
+RMSE: 4.640
+R2: 0.649
+Pearson: 0.805
+
+Metrics in training set
+MAE: 3.218
+RMSE: 6.308
+R2: 0.486
+Pearson: 0.700
+
+---- MLPRegressor ----
+
+Best model parameters: {'activation': 'relu', 'alpha': 0.001, 'hidden_layer_sizes': (50, 50), 'max_iter': 750}
+GridSearchCV fit: -4.5552258464380575
+
+Metrics in test set
+MAE: 3.255
+RMSE: 4.555
+R2: 0.661
+Pearson: 0.813
+
+Metrics in training set
+MAE: 2.864
+RMSE: 3.697
+R2: 0.823
+Pearson: 0.908
+
+---- XGBRegressor ----
+
+Best model parameters: {'learning_rate': 0.1, 'max_depth': 8, 'n_estimators': 50}
+GridSearchCV fit: -3.7466125602245017
+
+Metrics in test set
+MAE: 2.581
+RMSE: 3.747
+R2: 0.771
+Pearson: 0.878
+
+Metrics in training set
+MAE: 2.918
+RMSE: 4.633
+R2: 0.723
+Pearson: 0.862
+
+---- LinearRegression ----
+
+Best model parameters: {}
+GridSearchCV fit: -0.6723065259675224
+
+Metrics in test set
+MAE: 0.306
+RMSE: 0.672
+R2: 0.090
+Pearson: 0.301
+
+Metrics in training set
+MAE: 0.256
+RMSE: 0.545
+R2: 0.066
+Pearson: 0.276
+
+---- Lasso ----
+
+Best model parameters: {'alpha': 0.001}
+GridSearchCV fit: -0.6731771751574153
+
+Metrics in test set
+MAE: 0.306
+RMSE: 0.673
+R2: 0.088
+Pearson: 0.297
+
+Metrics in training set
+MAE: 0.251
+RMSE: 0.543
+R2: 0.073
+Pearson: 0.279
+
+---- Ridge ----
+
+Best model parameters: {'alpha': 0.05}
+GridSearchCV fit: -0.6724619131177046
+
+Metrics in test set
+MAE: 0.306
+RMSE: 0.672
+R2: 0.090
+Pearson: 0.300
+
+Metrics in training set
+MAE: 0.255
+RMSE: 0.544
+R2: 0.067
+Pearson: 0.276
+
+---- ElasticNet ----
+
+Best model parameters: {'alpha': 0.05, 'l1_ratio': 0.25}
+GridSearchCV fit: -0.6977921010953593
+
+Metrics in test set
+MAE: 0.334
+RMSE: 0.698
+R2: 0.020
+Pearson: 0.278
+
+Metrics in training set
+MAE: 0.259
+RMSE: 0.557
+R2: 0.024
+Pearson: 0.295
+
+---- DecisionTreeRegressor ----
+
+Best model parameters: {'max_depth': None}
+GridSearchCV fit: -0.6004098479616863
+
+Metrics in test set
+MAE: 0.257
+RMSE: 0.600
+R2: 0.275
+Pearson: 0.524
+
+Metrics in training set
+MAE: 0.235
+RMSE: 0.429
+R2: 0.420
+Pearson: 0.657
+
+---- RandomForestRegressor ----
+
+Best model parameters: {'max_depth': None, 'max_features': 1, 'min_samples_leaf': 1, 'min_samples_split': 4, 'n_estimators': 100}
+GridSearchCV fit: -0.6073093092838523
+
+Metrics in test set
+MAE: 0.267
+RMSE: 0.607
+R2: 0.258
+Pearson: 0.511
+
+Metrics in training set
+MAE: 0.244
+RMSE: 0.454
+R2: 0.352
+Pearson: 0.597
+
+---- GradientBoostingRegressor ----
+
+Best model parameters: {'learning_rate': 0.01, 'max_depth': 3, 'n_estimators': 150}
+GridSearchCV fit: -0.6353685552441649
+
+Metrics in test set
+MAE: 0.286
+RMSE: 0.635
+R2: 0.188
+Pearson: 0.465
+
+Metrics in training set
+MAE: 0.240
+RMSE: 0.506
+R2: 0.193
+Pearson: 0.474
+
+---- AdaBoostRegressor ----
+
+Best model parameters: {'learning_rate': 0.01, 'n_estimators': 100}
+GridSearchCV fit: -0.6493022586238576
+
+Metrics in test set
+MAE: 0.332
+RMSE: 0.649
+R2: 0.152
+Pearson: 0.403
+
+Metrics in training set
+MAE: 0.290
+RMSE: 0.540
+R2: 0.081
+Pearson: 0.345
+
+---- KNeighborsRegressor ----
+
+Best model parameters: {'n_neighbors': 10}
+GridSearchCV fit: -0.6543441572143739
+
+Metrics in test set
+MAE: 0.283
+RMSE: 0.654
+R2: 0.138
+Pearson: 0.382
+
+Metrics in training set
+MAE: 0.256
+RMSE: 0.547
+R2: 0.058
+Pearson: 0.277
+
+---- MLPRegressor ----
+
+Best model parameters: {'activation': 'tanh', 'alpha': 0.0005, 'hidden_layer_sizes': (25, 25), 'max_iter': 750}
+GridSearchCV fit: -0.6710469369980918
+
+Metrics in test set
+MAE: 0.317
+RMSE: 0.671
+R2: 0.094
+Pearson: 0.314
+
+Metrics in training set
+MAE: 0.267
+RMSE: 0.542
+R2: 0.074
+Pearson: 0.304
+
+---- XGBRegressor ----
+
+Best model parameters: {'learning_rate': 0.1, 'max_depth': 3, 'n_estimators': 25}
+GridSearchCV fit: -0.6308284408095344
+
+Metrics in test set
+MAE: 0.285
+RMSE: 0.631
+R2: 0.199
+Pearson: 0.465
+
+Metrics in training set
+MAE: 0.243
+RMSE: 0.491
+R2: 0.240
+Pearson: 0.510
+
+---- LinearRegression ----
+
+Best model parameters: {}
+GridSearchCV fit: -4.673054513199252
+
+Metrics in test set
+MAE: 3.395
+RMSE: 4.673
+R2: 0.644
+Pearson: 0.802
+
+Metrics in training set
+MAE: 2.940
+RMSE: 3.743
+R2: 0.819
+Pearson: 0.905
+
+---- Lasso ----
+
+Best model parameters: {'alpha': 0.001}
+GridSearchCV fit: -4.673851754917777
+
+Metrics in test set
+MAE: 3.397
+RMSE: 4.674
+R2: 0.644
+Pearson: 0.802
+
+Metrics in training set
+MAE: 2.948
+RMSE: 3.745
+R2: 0.819
+Pearson: 0.905
+
+---- Ridge ----
+
+Best model parameters: {'alpha': 0.01}
+GridSearchCV fit: -4.6732333376506965
+
+Metrics in test set
+MAE: 3.396
+RMSE: 4.673
+R2: 0.644
+Pearson: 0.802
+
+Metrics in training set
+MAE: 2.948
+RMSE: 3.745
+R2: 0.819
+Pearson: 0.905
+
+---- ElasticNet ----
+
+Best model parameters: {'alpha': 0.05, 'l1_ratio': 0.9}
+GridSearchCV fit: -5.36241111391284
+
+Metrics in test set
+MAE: 4.036
+RMSE: 5.362
+R2: 0.531
+Pearson: 0.758
+
+Metrics in training set
+MAE: 3.825
+RMSE: 6.789
+R2: 0.404
+Pearson: 0.669
+
+---- DecisionTreeRegressor ----
+
+Best model parameters: {'max_depth': 10}
+GridSearchCV fit: -3.7898251391858016
+
+Metrics in test set
+MAE: 2.601
+RMSE: 3.790
+R2: 0.766
+Pearson: 0.875
+
+Metrics in training set
+MAE: 3.012
+RMSE: 4.740
+R2: 0.710
+Pearson: 0.849
+
+---- RandomForestRegressor ----
+
+Best model parameters: {'max_depth': None, 'max_features': 1, 'min_samples_leaf': 1, 'min_samples_split': 2, 'n_estimators': 50}
+GridSearchCV fit: -3.8182458821700864
+
+Metrics in test set
+MAE: 2.653
+RMSE: 3.818
+R2: 0.762
+Pearson: 0.873
+
+Metrics in training set
+MAE: 2.871
+RMSE: 4.872
+R2: 0.693
+Pearson: 0.847
+
+---- GradientBoostingRegressor ----
+
+Best model parameters: {'learning_rate': 0.1, 'max_depth': 5, 'n_estimators': 50}
+GridSearchCV fit: -3.7839595159706927
+
+Metrics in test set
+MAE: 2.638
+RMSE: 3.784
+R2: 0.766
+Pearson: 0.876
+
+Metrics in training set
+MAE: 2.806
+RMSE: 4.458
+R2: 0.743
+Pearson: 0.875
+
+---- AdaBoostRegressor ----
+
+Best model parameters: {'learning_rate': 0.05, 'n_estimators': 50}
+GridSearchCV fit: -4.503353839005145
+
+Metrics in test set
+MAE: 3.305
+RMSE: 4.503
+R2: 0.669
+Pearson: 0.818
+
+Metrics in training set
+MAE: 3.093
+RMSE: 4.718
+R2: 0.712
+Pearson: 0.854
+
+---- KNeighborsRegressor ----
+
+Best model parameters: {'n_neighbors': 10}
+GridSearchCV fit: -4.640041949298464
+
+Metrics in test set
+MAE: 3.142
+RMSE: 4.640
+R2: 0.649
+Pearson: 0.805
+
+Metrics in training set
+MAE: 3.218
+RMSE: 6.308
+R2: 0.486
+Pearson: 0.700
+
+---- MLPRegressor ----
+
+Best model parameters: {'activation': 'relu', 'alpha': 0.0001, 'hidden_layer_sizes': (50, 50), 'max_iter': 750}
+GridSearchCV fit: -4.548084675098752
+
+Metrics in test set
+MAE: 3.251
+RMSE: 4.548
+R2: 0.662
+Pearson: 0.814
+
+Metrics in training set
+MAE: 2.828
+RMSE: 3.639
+R2: 0.829
+Pearson: 0.911
+
+---- XGBRegressor ----
+
+Best model parameters: {'learning_rate': 0.1, 'max_depth': 8, 'n_estimators': 50}
+GridSearchCV fit: -3.7466125602245017
+
+Metrics in test set
+MAE: 2.581
+RMSE: 3.747
+R2: 0.771
+Pearson: 0.878
+
+Metrics in training set
+MAE: 2.918
+RMSE: 4.633
+R2: 0.723
+Pearson: 0.862
+
+---- LinearRegression ----
+
+Best model parameters: {}
+GridSearchCV fit: -4.673054513199252
+
+Metrics in test set
+MAE: 3.395
+RMSE: 4.673
+R2: 0.644
+Pearson: 0.802
+
+Metrics in training set
+MAE: 2.940
+RMSE: 3.743
+R2: 0.819
+Pearson: 0.905
+
+---- Lasso ----
+
+Best model parameters: {'alpha': 0.001}
+GridSearchCV fit: -4.673851754917777
+
+Metrics in test set
+MAE: 3.397
+RMSE: 4.674
+R2: 0.644
+Pearson: 0.802
+
+Metrics in training set
+MAE: 2.948
+RMSE: 3.745
+R2: 0.819
+Pearson: 0.905
+
+---- Ridge ----
+
+Best model parameters: {'alpha': 0.01}
+GridSearchCV fit: -4.6732333376506965
+
+Metrics in test set
+MAE: 3.396
+RMSE: 4.673
+R2: 0.644
+Pearson: 0.802
+
+Metrics in training set
+MAE: 2.948
+RMSE: 3.745
+R2: 0.819
+Pearson: 0.905
+
+---- ElasticNet ----
+
+Best model parameters: {'alpha': 0.05, 'l1_ratio': 0.9}
+GridSearchCV fit: -5.36241111391284
+
+Metrics in test set
+MAE: 4.036
+RMSE: 5.362
+R2: 0.531
+Pearson: 0.758
+
+Metrics in training set
+MAE: 3.825
+RMSE: 6.789
+R2: 0.404
+Pearson: 0.669
+
+---- DecisionTreeRegressor ----
+
+Best model parameters: {'max_depth': 10}
+GridSearchCV fit: -3.7898251391858016
+
+Metrics in test set
+MAE: 2.601
+RMSE: 3.790
+R2: 0.766
+Pearson: 0.875
+
+Metrics in training set
+MAE: 3.009
+RMSE: 4.739
+R2: 0.710
+Pearson: 0.849
+
+---- RandomForestRegressor ----
+
+Best model parameters: {'max_depth': None, 'max_features': 1, 'min_samples_leaf': 1, 'min_samples_split': 2, 'n_estimators': 100}
+GridSearchCV fit: -3.8151132180564558
+
+Metrics in test set
+MAE: 2.654
+RMSE: 3.815
+R2: 0.763
+Pearson: 0.874
+
+Metrics in training set
+MAE: 2.870
+RMSE: 5.023
+R2: 0.674
+Pearson: 0.834
+
+---- GradientBoostingRegressor ----
+
+Best model parameters: {'learning_rate': 0.5, 'max_depth': 3, 'n_estimators': 25}
+GridSearchCV fit: -3.7963663576274578
+
+Metrics in test set
+MAE: 2.655
+RMSE: 3.796
+R2: 0.765
+Pearson: 0.875
+
+Metrics in training set
+MAE: 2.915
+RMSE: 4.522
+R2: 0.736
+Pearson: 0.866
+
+---- AdaBoostRegressor ----
+
+Best model parameters: {'learning_rate': 0.5, 'n_estimators': 100}
+GridSearchCV fit: -4.422867684560212
+
+Metrics in test set
+MAE: 3.252
+RMSE: 4.423
+R2: 0.681
+Pearson: 0.825
+
+Metrics in training set
+MAE: 3.200
+RMSE: 4.928
+R2: 0.686
+Pearson: 0.839
+
+---- KNeighborsRegressor ----
+
+Best model parameters: {'n_neighbors': 10}
+GridSearchCV fit: -4.640041949298464
+
+Metrics in test set
+MAE: 3.142
+RMSE: 4.640
+R2: 0.649
+Pearson: 0.805
+
+Metrics in training set
+MAE: 3.218
+RMSE: 6.308
+R2: 0.486
+Pearson: 0.700
+
+---- MLPRegressor ----
+
+Best model parameters: {'activation': 'relu', 'alpha': 0.0005, 'hidden_layer_sizes': (50, 50), 'max_iter': 750}
+GridSearchCV fit: -4.548041912359535
+
+Metrics in test set
+MAE: 3.267
+RMSE: 4.548
+R2: 0.662
+Pearson: 0.814
+
+Metrics in training set
+MAE: 2.797
+RMSE: 3.590
+R2: 0.833
+Pearson: 0.914
+
+---- XGBRegressor ----
+
+Best model parameters: {'learning_rate': 0.1, 'max_depth': 8, 'n_estimators': 50}
+GridSearchCV fit: -3.7466125602245017
+
+Metrics in test set
+MAE: 2.581
+RMSE: 3.747
+R2: 0.771
+Pearson: 0.878
+
+Metrics in training set
+MAE: 2.918
+RMSE: 4.633
+R2: 0.723
+Pearson: 0.862
+
+---- LinearRegression ----
+
+Best model parameters: {}
+GridSearchCV fit: -4.510620707568448
+
+Metrics in test set
+MAE: 3.305
+RMSE: 4.511
+R2: 0.704
+Pearson: 0.839
+
+Metrics in training set
+MAE: 3.333
+RMSE: 4.419
+R2: 0.577
+Pearson: 0.763
+
+---- Lasso ----
+
+Best model parameters: {'alpha': 0.001}
+GridSearchCV fit: -4.51237670165775
+
+Metrics in test set
+MAE: 3.308
+RMSE: 4.512
+R2: 0.704
+Pearson: 0.839
+
+Metrics in training set
+MAE: 3.327
+RMSE: 4.416
+R2: 0.577
+Pearson: 0.764
+
+---- Ridge ----
+
+Best model parameters: {'alpha': 0.01}
+GridSearchCV fit: -4.510811201931662
+
+Metrics in test set
+MAE: 3.305
+RMSE: 4.511
+R2: 0.704
+Pearson: 0.839
+
+Metrics in training set
+MAE: 3.332
+RMSE: 4.417
+R2: 0.577
+Pearson: 0.764
+
+---- ElasticNet ----
+
+Best model parameters: {'alpha': 0.05, 'l1_ratio': 0.9}
+GridSearchCV fit: -5.470696893464987
+
+Metrics in test set
+MAE: 3.972
+RMSE: 5.471
+R2: 0.565
+Pearson: 0.791
+
+Metrics in training set
+MAE: 3.769
+RMSE: 4.740
+R2: 0.513
+Pearson: 0.752
+
+---- DecisionTreeRegressor ----
+
+Best model parameters: {'max_depth': 10}
+GridSearchCV fit: -3.73702098005142
+
+Metrics in test set
+MAE: 2.553
+RMSE: 3.737
+R2: 0.797
+Pearson: 0.893
+
+Metrics in training set
+MAE: 2.967
+RMSE: 4.291
+R2: 0.601
+Pearson: 0.777
+
+---- RandomForestRegressor ----
+
+Best model parameters: {'max_depth': None, 'max_features': 1, 'min_samples_leaf': 1, 'min_samples_split': 2, 'n_estimators': 25}
+GridSearchCV fit: -3.7842360585440136
+
+Metrics in test set
+MAE: 2.625
+RMSE: 3.784
+R2: 0.792
+Pearson: 0.891
+
+Metrics in training set
+MAE: 2.732
+RMSE: 3.798
+R2: 0.687
+Pearson: 0.829
+
+---- GradientBoostingRegressor ----
+
+Best model parameters: {'learning_rate': 0.1, 'max_depth': 7, 'n_estimators': 25}
+GridSearchCV fit: -3.7884139588151604
+
+Metrics in test set
+MAE: 2.733
+RMSE: 3.788
+R2: 0.791
+Pearson: 0.892
+
+Metrics in training set
+MAE: 3.007
+RMSE: 4.172
+R2: 0.622
+Pearson: 0.791
+
+---- AdaBoostRegressor ----
+
+Best model parameters: {'learning_rate': 0.01, 'n_estimators': 200}
+GridSearchCV fit: -4.3629534092738025
+
+Metrics in test set
+MAE: 3.219
+RMSE: 4.363
+R2: 0.723
+Pearson: 0.851
+
+Metrics in training set
+MAE: 3.400
+RMSE: 4.483
+R2: 0.564
+Pearson: 0.753
+
+---- KNeighborsRegressor ----
+
+Best model parameters: {'n_neighbors': 7}
+GridSearchCV fit: -4.981489426837481
+
+Metrics in test set
+MAE: 3.155
+RMSE: 4.981
+R2: 0.639
+Pearson: 0.801
+
+Metrics in training set
+MAE: 2.965
+RMSE: 3.906
+R2: 0.669
+Pearson: 0.822
+
+---- MLPRegressor ----
+
+Best model parameters: {'activation': 'relu', 'alpha': 0.001, 'hidden_layer_sizes': (50, 50), 'max_iter': 750}
+GridSearchCV fit: -4.313999868675119
+
+Metrics in test set
+MAE: 3.044
+RMSE: 4.314
+R2: 0.729
+Pearson: 0.854
+
+Metrics in training set
+MAE: 3.155
+RMSE: 4.246
+R2: 0.609
+Pearson: 0.782
+
+---- XGBRegressor ----
+
+Best model parameters: {'learning_rate': 0.25, 'max_depth': 5, 'n_estimators': 25}
+GridSearchCV fit: -3.7959871302854866
+
+Metrics in test set
+MAE: 2.676
+RMSE: 3.796
+R2: 0.790
+Pearson: 0.889
+
+Metrics in training set
+MAE: 2.885
+RMSE: 3.943
+R2: 0.663
+Pearson: 0.814
+
diff --git a/scratch/checkContentTST.py b/scratch/checkContentTST.py
new file mode 100644
index 00000000..82657c98
--- /dev/null
+++ b/scratch/checkContentTST.py
@@ -0,0 +1,137 @@
+def getTotalNlipids(system):
+ NLIPIDS = 0
+ for molecule in system['COMPOSITION']:
+ if molecule in lipids_dict:
+ NLIPIDS += np.sum(system['COMPOSITION'][molecule]['COUNT'])
+ return NLIPIDS
+
+def getTotalNsolvent(system):
+ NMOLECULES = 0
+ for molecule in system['COMPOSITION']:
+ if molecule not in lipids_dict:
+ NMOLECULES += np.sum(system['COMPOSITION'][molecule]['COUNT'])
+ return NMOLECULES
+
+def checkAvailabilitySIM(system,lipid,counterion):
+ status = {}
+ TotalNlipids = getTotalNlipids(system)
+ TotalNsolvent = getTotalNsolvent(system)
+ Nwater = system['COMPOSITION']['SOL']['COUNT']
+ if counterion != 'no':
+ try:
+ Ncounterion = system['COMPOSITION'][counterion]['COUNT']
+ except:
+ Ncounterion = 0
+
+ try:
+ Nlipid = np.sum(system['COMPOSITION'][lipid]['COUNT'])
+ except:
+ Nlipid = 0
+
+ path = system['path']
+
+ QualityEvaluated = False
+ TotalQualityFilePath = path + '/SYSTEM_quality.json'
+ if (os.path.isfile(TotalQualityFilePath)):
+ with open(TotalQualityFilePath) as json_file:
+ Quality = json.load(json_file)
+ json_file.close()
+ if all(value > 0 for value in Quality.values()):
+ #print(Quality)
+ QualityEvaluated = True
+
+ xrayQualityEvaluated = False
+ xrayQualityFilePath = path + '/FormFactorQuality.json'
+ if (os.path.isfile(xrayQualityFilePath)):
+ with open(xrayQualityFilePath) as json_file:
+ xrayQuality = json.load(json_file)
+ json_file.close()
+ if len(xrayQuality) > 0 and xrayQuality[0] > 0:
+ xrayQualityEvaluated = True
+
+ SingleComponentSystem = False
+ if Nlipid == TotalNlipids:
+ if counterion == 'no' and Nwater == TotalNsolvent:
+ SingleComponentSystem = True
+ if counterion != 'no'and Nwater == TotalNsolvent-Ncounterion and Nlipid == Ncounterion:
+ SingleComponentSystem = True
+
+ if SingleComponentSystem:
+ status['Simulation'] = 'yes'
+ status['FF'] = system['FF']
+ if QualityEvaluated:
+ status['Experiment'] = 'yes'
+ status['Quality'] = Quality['total']
+
+ if not QualityEvaluated:
+ status['Experiment'] = 'no'
+ status['Quality'] = 0
+
+ if xrayQualityEvaluated:
+ status['xrayExperiment'] = 'yes'
+ status['xrayQuality'] = xrayQuality[0]
+
+ if not xrayQualityEvaluated:
+ status['xrayExperiment'] = 'no'
+ status['xrayQuality'] = 0
+
+ return status
+
+
+def giveStatus(systems,lipid,counterion):
+ status = {'Simulation': 'no', 'Experiment': 'no', 'Quality': 0, 'xrayExperiment': 'no', 'xrayQuality': 0}
+ QualityEvaluatedFound = False
+ for system in systems:
+ TMPstatus = checkAvailabilitySIM(system,lipid,counterion)
+ if TMPstatus and TMPstatus['Quality'] > status['Quality']:
+ QualityEvaluatedFound = True
+ status = TMPstatus
+ if TMPstatus and not QualityEvaluatedFound and TMPstatus['Simulation'] == 'yes':
+ status = TMPstatus
+ return status
+
+def giveExpStatus(lipids,counterion,status):
+
+
+
+ return status
+
+
+
+HGs = {'PC', 'PE', 'PG', 'PS'}
+tails = {'PO', 'DO', 'DP'}
+table = {}
+
+for tail in tails:
+ table[tail] = {}
+ for HG in HGs:
+ lipid = tail + HG
+
+ if lipid == 'POPS' or lipid == 'POPG' or lipid == 'DPPG':
+ counterion = 'SOD'
+ else:
+ counterion = 'no'
+
+ status = giveStatus(systems,lipid,counterion)
+ statusString = ''
+
+ if status['Quality'] > 0:
+ if 'ECC-lipids' in status['FF']:
+ FF = 'ECClipids'
+ else:
+ FF = status['FF']
+ #print(status['xrayQuality'])
+ statusString = FF + '(' + str(round(status['Quality'],2)) + ',' + str(round(status['xrayQuality'])) + ')'
+ else:
+ if status['Simulation'] == 'no':
+ statusString = statusString + 'MD,'
+ if status['Experiment'] == 'no':
+ statusString = statusString + 'NMR,'
+ if status['xrayExperiment'] == 'no':
+ statusString = statusString + 'x-ray'
+
+ table[tail][HG] = statusString
+
+
+print(pd.DataFrame(table))
+display(pd.DataFrame(table))
diff --git a/scripts/APLpredictor.ipynb b/scripts/APLpredictor.ipynb
index b67052aa..4476ff0d 100644
--- a/scripts/APLpredictor.ipynb
+++ b/scripts/APLpredictor.ipynb
@@ -1,5 +1,23 @@
{
"cells": [
+ {
+ "cell_type": "markdown",
+ "id": "b881194c",
+ "metadata": {},
+ "source": [
+ "### You can run this notebook at Colab by clicking here:"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "7ac89eed",
+ "metadata": {},
+ "source": [
+ "\n",
+ " \n",
+ ""
+ ]
+ },
{
"cell_type": "markdown",
"id": "22724830",
@@ -22,7 +40,7 @@
},
{
"cell_type": "code",
- "execution_count": 1,
+ "execution_count": 25,
"id": "70d500f2",
"metadata": {},
"outputs": [
@@ -40,18 +58,27 @@
"import numpy as np\n",
"import json\n",
"import matplotlib.pyplot as plt\n",
- "import MDAnalysis\n",
"import urllib.request\n",
"import yaml\n",
"import random\n",
"import collections\n",
"import pandas as pd\n",
"\n",
+ "if 'google.colab' in sys.modules:\n",
+ " !pip3 install MDAnalysis\n",
+ "import MDAnalysis\n",
+ "\n",
+ "\n",
"# This defines the path for the NMRlipids databank on your computer. \n",
"# Default is that this repository and the NMRlipids databank repository are cloned to the same folder.\n",
"# If this is not the case, change this to the folder where the NMRlipids databank repository is located.\n",
+ "\n",
"databankPath = '../../Databank/'\n",
"\n",
+ "if 'google.colab' in sys.modules:\n",
+ " !git clone https://github.com/NMRLipids/Databank.git\n",
+ " databankPath = '/content/Databank'\n",
+ "\n",
"# This enales the access to functions defined in the NMRlipids databank.\n",
"sys.path.insert(1, databankPath + '/Scripts/BuildDatabank/')\n",
"from databankLibrary import * \n",
@@ -63,7 +90,7 @@
},
{
"cell_type": "code",
- "execution_count": 2,
+ "execution_count": 26,
"id": "b94f1394",
"metadata": {},
"outputs": [],
@@ -248,7 +275,7 @@
},
{
"cell_type": "code",
- "execution_count": 3,
+ "execution_count": 27,
"id": "9c5dea7f",
"metadata": {},
"outputs": [
@@ -256,7 +283,7 @@
"name": "stdout",
"output_type": "stream",
"text": [
- "{'POPC': 480, 'CHOL': 189, 'DMPC': 15, 'DMTAP': 9, 'DPPC': 73, 'DOPS': 13, 'POPE': 59, 'DOPC': 38, 'POPS': 86, 'POPG': 63, 'DRPC': 3, 'DAPC': 15, 'DOPE': 8, 'SDG': 4, 'SAPI24': 1, 'SM16': 5, 'CER': 5, 'DLIPC': 6, 'TLCL_0H': 3, 'SM18': 2, 'PYPC': 3, 'GM1': 3, 'DLPC': 8, 'DHMDMAB': 13, 'DOG': 4, 'POPI': 2, 'CER180': 1, 'DYPC': 3, 'TOCL': 5, 'SOPC': 6, 'SAPI25': 6, 'DSPC': 2, 'SLiPC': 3, 'SAPI': 1, 'SLPI': 1, 'DEPC': 1, 'PAzePCprot': 1, 'DPPGK': 1, 'SDPE': 4, 'DDOPC': 2, 'PAzePCdeprot': 1, 'DPPG': 2, 'DPPE': 3, 'DCHOL': 2}\n",
+ "{'POPC': 481, 'CHOL': 189, 'DMPC': 17, 'DMTAP': 9, 'DPPC': 73, 'DOPS': 13, 'POPE': 61, 'DOPC': 38, 'POPS': 86, 'POPG': 63, 'DRPC': 3, 'DAPC': 15, 'DOPE': 8, 'SDG': 4, 'SAPI24': 1, 'SM16': 5, 'CER': 5, 'DLIPC': 6, 'TLCL_0H': 3, 'SM18': 2, 'PYPC': 3, 'GM1': 3, 'DLPC': 8, 'DHMDMAB': 13, 'DOG': 4, 'POPI': 2, 'CER180': 1, 'DYPC': 3, 'TOCL': 5, 'SOPC': 6, 'SAPI25': 6, 'DSPC': 2, 'SLiPC': 3, 'SAPI': 1, 'SLPI': 1, 'DEPC': 1, 'PAzePCprot': 1, 'DPPGK': 1, 'SDPE': 4, 'DDOPC': 2, 'PAzePCdeprot': 1, 'DPPG': 2, 'DPPE': 3, 'DCHOL': 2}\n",
"['POPC', 'CHOL', 'DMPC', 'DMTAP', 'DPPC', 'DOPS', 'POPE', 'DOPC', 'POPS', 'POPG', 'DRPC', 'DAPC', 'DOPE', 'SDG', 'SAPI24', 'SM16', 'CER', 'DLIPC', 'TLCL_0H', 'SM18', 'PYPC', 'GM1', 'DLPC', 'DHMDMAB', 'DOG', 'POPI', 'CER180', 'DYPC', 'TOCL', 'SOPC', 'SAPI25', 'DSPC', 'SLiPC', 'SAPI', 'SLPI', 'DEPC', 'PAzePCprot', 'DPPGK', 'SDPE', 'DDOPC', 'PAzePCdeprot', 'DPPG', 'DPPE', 'DCHOL']\n"
]
}
@@ -341,7 +368,7 @@
"name": "stdout",
"output_type": "stream",
"text": [
- "765\n"
+ "770\n"
]
}
],
@@ -540,12 +567,12 @@
"name": "stdout",
"output_type": "stream",
"text": [
- "LinearRegression was finalized in 0.7970 seconds\n",
- "Lasso was finalized in 0.0286 seconds\n",
- "Ridge was finalized in 0.0262 seconds\n",
- "ElasticNet was finalized in 0.0678 seconds\n",
- "DecisionTreeRegressor was finalized in 0.0436 seconds\n",
- "RandomForestRegressor was finalized in 39.8423 seconds\n"
+ "LinearRegression was finalized in 0.8348 seconds\n",
+ "Lasso was finalized in 0.0376 seconds\n",
+ "Ridge was finalized in 0.0274 seconds\n",
+ "ElasticNet was finalized in 0.0667 seconds\n",
+ "DecisionTreeRegressor was finalized in 0.0409 seconds\n",
+ "RandomForestRegressor was finalized in 38.5872 seconds\n"
]
},
{
@@ -559,7 +586,7 @@
"\n",
"Below are more details about the failures:\n",
"--------------------------------------------------------------------------------\n",
- "1045 fits failed with the following error:\n",
+ "805 fits failed with the following error:\n",
"Traceback (most recent call last):\n",
" File \"/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 729, in _fit_and_score\n",
" estimator.fit(X_train, y_train, **fit_params)\n",
@@ -569,10 +596,10 @@
" validate_parameter_constraints(\n",
" File \"/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/utils/_param_validation.py\", line 95, in validate_parameter_constraints\n",
" raise InvalidParameterError(\n",
- "sklearn.utils._param_validation.InvalidParameterError: The 'max_features' parameter of RandomForestRegressor must be an int in the range [1, inf), a float in the range (0.0, 1.0], a str among {'log2', 'sqrt'} or None. Got 'auto' instead.\n",
+ "sklearn.utils._param_validation.InvalidParameterError: The 'max_features' parameter of RandomForestRegressor must be an int in the range [1, inf), a float in the range (0.0, 1.0], a str among {'sqrt', 'log2'} or None. Got 'auto' instead.\n",
"\n",
"--------------------------------------------------------------------------------\n",
- "955 fits failed with the following error:\n",
+ "1195 fits failed with the following error:\n",
"Traceback (most recent call last):\n",
" File \"/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/model_selection/_validation.py\", line 729, in _fit_and_score\n",
" estimator.fit(X_train, y_train, **fit_params)\n",
@@ -582,7 +609,7 @@
" validate_parameter_constraints(\n",
" File \"/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/utils/_param_validation.py\", line 95, in validate_parameter_constraints\n",
" raise InvalidParameterError(\n",
- "sklearn.utils._param_validation.InvalidParameterError: The 'max_features' parameter of RandomForestRegressor must be an int in the range [1, inf), a float in the range (0.0, 1.0], a str among {'sqrt', 'log2'} or None. Got 'auto' instead.\n",
+ "sklearn.utils._param_validation.InvalidParameterError: The 'max_features' parameter of RandomForestRegressor must be an int in the range [1, inf), a float in the range (0.0, 1.0], a str among {'log2', 'sqrt'} or None. Got 'auto' instead.\n",
"\n",
" warnings.warn(some_fits_failed_message, FitFailedWarning)\n",
"/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/model_selection/_search.py:979: UserWarning: One or more of the test scores are non-finite: [ nan nan nan nan nan nan\n",
@@ -598,20 +625,20 @@
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
- " nan nan -4.64114821 -4.61852854 -4.63117993 -4.61662176\n",
- " -4.69274686 -4.62725027 -4.65503904 -4.68020476 -4.60157233 -4.70954703\n",
- " -4.67619448 -4.68780436 -4.80861558 -4.75457724 -4.71608726 -4.71660454\n",
- " -4.67400562 -4.70711643 -4.70906214 -4.69687263 -5.55220044 -5.21830929\n",
- " -5.18834153 -5.31225837 -5.18172406 -5.12348193 -5.21595419 -5.23373727\n",
- " -5.42478906 -5.25423608 -5.30207757 -5.26483871 -5.50458958 -5.2179175\n",
- " -5.2189786 -5.26530203 -5.31605834 -5.25591403 -5.21021107 -5.15162523\n",
- " -6.37758625 -6.35686986 -6.34755593 -6.39946903 -6.35920665 -6.45304259\n",
- " -6.41494327 -6.34311767 -6.53246956 -6.48804797 -6.46884464 -6.37769022\n",
- " -6.22354246 -6.6405786 -6.50315734 -6.39151569 -6.62871432 -6.52625912\n",
- " -6.46871898 -6.37256801 -6.60968902 -6.56930381 -6.69968877 -6.58504114\n",
- " -6.72387417 -6.5684446 -6.59488101 -6.58912697 -6.83864387 -6.73932422\n",
- " -6.54725877 -6.63690271 -6.67019023 -6.63092191 -6.75350958 -6.62511791\n",
- " -6.56631303 -6.57092397 -6.63011462 -6.66415308 nan nan\n",
+ " nan nan -4.73842843 -4.78228161 -4.76743917 -4.78375469\n",
+ " -4.90197845 -4.84701453 -4.83408902 -4.83481387 -4.87047111 -4.81269746\n",
+ " -4.81415078 -4.84605643 -4.92360274 -4.82938721 -4.81813365 -4.83731581\n",
+ " -4.94596317 -4.79661341 -4.79486795 -4.78390189 -5.17071995 -5.03771754\n",
+ " -5.16521701 -5.13248472 -5.27859797 -5.09831483 -5.12746111 -5.16251411\n",
+ " -5.16991057 -5.22980494 -5.18834004 -5.24160981 -5.20102507 -5.2902974\n",
+ " -5.16901257 -5.16078109 -5.09299137 -5.33111432 -5.17074785 -5.1728084\n",
+ " -6.39398965 -6.41274839 -6.47069012 -6.39583678 -6.4819712 -6.46122747\n",
+ " -6.45408373 -6.43569546 -6.16846585 -6.44661799 -6.45073687 -6.42810308\n",
+ " -7.00658972 -6.404748 -6.50904736 -6.30692934 -6.48899594 -6.53993301\n",
+ " -6.461672 -6.41050542 -7.03972134 -6.9444264 -6.98210352 -7.17175374\n",
+ " -7.10317353 -7.17693146 -7.12908468 -7.10176885 -6.99734011 -7.09313053\n",
+ " -7.11193374 -7.03303462 -6.99023901 -7.0992357 -7.12507375 -7.09563889\n",
+ " -7.26790786 -7.11749793 -7.04680214 -7.1414514 nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
@@ -625,20 +652,20 @@
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
- " -6.09561559 -6.20561877 -5.98203605 -6.10053335 -5.98226221 -6.07851053\n",
- " -6.12487238 -6.09553075 -6.18621511 -6.20191212 -6.05588093 -6.14776924\n",
- " -5.97666265 -6.16462596 -6.12721955 -6.09441958 -6.23010996 -6.16501269\n",
- " -6.11069029 -6.14363454 -6.28821037 -6.24597575 -6.29311171 -6.2782975\n",
- " -6.24995106 -6.10893246 -6.3627363 -6.25089086 -6.22150422 -6.21700306\n",
- " -6.31851402 -6.28353839 -6.18343303 -6.22726082 -6.26463388 -6.12197901\n",
- " -6.07688678 -6.17723257 -6.39122815 -6.10612223 -6.65649445 -6.66553462\n",
- " -6.62516566 -6.62869251 -6.64342825 -6.5915944 -6.62363717 -6.6025837\n",
- " -6.45819064 -6.5094724 -6.51535618 -6.74750504 -6.53878773 -6.60860811\n",
- " -6.67439564 -6.56729864 -6.72941917 -6.7602782 -6.62860393 -6.62579828\n",
- " -6.71668758 -6.83703182 -6.76613035 -6.7975 -7.04386227 -6.72856555\n",
- " -6.83989467 -6.7649139 -6.90955992 -6.9415778 -6.8198139 -6.81716601\n",
- " -6.76913848 -6.82420748 -6.88468237 -6.74467515 -6.7113096 -6.88262011\n",
- " -6.78452741 -6.83222654 nan nan nan nan\n",
+ " -6.21382386 -6.45822812 -6.41459987 -6.50217626 -6.51054539 -6.45094412\n",
+ " -6.45807365 -6.40554994 -6.38317265 -6.56472464 -6.46683341 -6.41675531\n",
+ " -6.44700173 -6.52945192 -6.36243113 -6.53492684 -6.4021603 -6.50715076\n",
+ " -6.49248965 -6.43098656 -6.44430164 -6.31252505 -6.42964145 -6.47396933\n",
+ " -6.32977169 -6.53431822 -6.37335648 -6.49004668 -6.41609723 -6.45505936\n",
+ " -6.40457796 -6.38827721 -6.42991797 -6.39011717 -6.45836642 -6.36522444\n",
+ " -6.64394839 -6.47344241 -6.41062352 -6.41242534 -6.86252159 -6.85093731\n",
+ " -6.81215684 -6.87079048 -6.73700144 -6.75105411 -6.82741318 -6.7655989\n",
+ " -7.01759065 -6.75185219 -6.91610096 -6.80657466 -6.81310658 -6.85174343\n",
+ " -6.79743294 -6.80914124 -6.94161436 -6.7677494 -6.77920507 -6.78629805\n",
+ " -6.84967255 -7.22058035 -7.22109734 -7.32200141 -7.22039069 -7.05494164\n",
+ " -7.26479333 -7.15132023 -7.2705475 -7.10756631 -7.0651567 -7.15274622\n",
+ " -7.10949795 -7.36765857 -7.18765026 -7.17866167 -7.1113454 -7.17175644\n",
+ " -7.14682041 -7.21990618 nan nan nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
@@ -651,20 +678,20 @@
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
- " nan nan nan nan -5.42751827 -5.42552641\n",
- " -5.54728766 -5.39481049 -5.43693692 -5.38577902 -5.29122598 -5.40324673\n",
- " -5.41903239 -5.2963599 -5.46111437 -5.37454816 -5.52577318 -5.40541344\n",
- " -5.48756159 -5.31502097 -5.4080328 -5.32816013 -5.47252423 -5.39513765\n",
- " -5.75992141 -5.7716831 -5.77194614 -5.71939354 -5.83672306 -5.80465235\n",
- " -5.68604585 -5.79634893 -5.62439133 -5.6774473 -5.68374577 -5.70158351\n",
- " -5.6545861 -5.65044789 -5.65936873 -5.70276098 -5.7353863 -5.78126823\n",
- " -5.63404803 -5.59413252 -6.44815354 -6.513721 -6.49073497 -6.39296496\n",
- " -6.56768142 -6.43141266 -6.34286855 -6.50731874 -6.56810368 -6.16300496\n",
- " -6.42567747 -6.47885866 -6.50560105 -6.6435576 -6.4316358 -6.28979541\n",
- " -6.57758484 -6.51726112 -6.52939184 -6.54896747 -6.7128315 -6.45781498\n",
- " -6.65983317 -6.64379268 -6.59611973 -6.5516572 -6.64131878 -6.62230184\n",
- " -6.86271713 -6.62523146 -6.67059831 -6.63975977 -6.63688613 -6.55626578\n",
- " -6.61658817 -6.68483708 -6.62125564 -6.68410706 -6.68049529 -6.69939221\n",
+ " nan nan nan nan -5.68705407 -5.60700815\n",
+ " -5.69516637 -5.59760156 -5.74185826 -5.67009394 -5.67768116 -5.62605548\n",
+ " -5.68492156 -5.51004721 -5.72303527 -5.65763714 -5.70629893 -5.7080233\n",
+ " -5.60066048 -5.66614897 -5.78818724 -5.65980379 -5.63765169 -5.70120101\n",
+ " -5.83209985 -5.74256745 -5.76937087 -5.76394646 -5.84290528 -5.72914408\n",
+ " -5.78645766 -5.75158699 -5.77118838 -5.76414796 -5.79651439 -5.89389357\n",
+ " -5.77361072 -5.84851592 -5.742141 -5.79996261 -5.92384025 -5.7696537\n",
+ " -5.72621872 -5.78141963 -6.51757104 -6.52527057 -6.56309712 -6.5476236\n",
+ " -6.33552411 -6.66362909 -6.60777147 -6.54732507 -6.4630971 -6.64984598\n",
+ " -6.66049802 -6.59577057 -6.61923316 -6.52694717 -6.5347218 -6.55035718\n",
+ " -6.41417388 -6.53631404 -6.51232736 -6.54604939 -7.15512415 -7.25371604\n",
+ " -7.19711076 -6.98089593 -6.93796647 -7.14481249 -7.0028551 -7.01277866\n",
+ " -6.86446533 -7.18138378 -7.0469604 -7.0917255 -7.02059149 -7.19709806\n",
+ " -7.10929809 -7.09757588 -7.03496533 -7.01630988 -7.14403744 -7.02561758\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
@@ -678,20 +705,20 @@
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
- " nan nan -5.22288049 -5.14673717 -5.12948563 -5.11461827\n",
- " -5.19802574 -5.05094638 -5.11326981 -5.16650906 -5.07509659 -5.07436246\n",
- " -5.22411137 -5.1294311 -5.14311676 -5.18995735 -5.06277713 -5.15754713\n",
- " -5.12209037 -5.15701192 -5.14835036 -5.18011592 -5.44004382 -5.48640514\n",
- " -5.39835862 -5.60063829 -5.56978708 -5.5486566 -5.52804806 -5.53258645\n",
- " -5.67972077 -5.61009482 -5.53298062 -5.49883561 -5.66165345 -5.43805104\n",
- " -5.50924027 -5.57518076 -5.55788842 -5.6165555 -5.56047351 -5.54795823\n",
- " -6.52636962 -6.52434947 -6.47697241 -6.42131988 -6.33193184 -6.46541689\n",
- " -6.52914047 -6.36447712 -6.36959748 -6.4788886 -6.37764679 -6.34867014\n",
- " -6.40090063 -6.28353913 -6.42476167 -6.52613103 -6.58765952 -6.48126248\n",
- " -6.39462401 -6.53498534 -6.82685502 -6.61335055 -6.58223568 -6.68771862\n",
- " -6.67026619 -6.65661887 -6.56763389 -6.6836883 -6.84764822 -6.8460395\n",
- " -6.58465816 -6.5923604 -6.52313467 -6.73860035 -6.63727092 -6.7254683\n",
- " -6.56444003 -6.71202555 -6.55940918 -6.64777249 nan nan\n",
+ " nan nan -5.39738938 -5.38079106 -5.3923018 -5.3060259\n",
+ " -5.5380233 -5.36829461 -5.31206218 -5.36676445 -5.4305634 -5.41418515\n",
+ " -5.3995325 -5.41625083 -5.51233848 -5.32806396 -5.37604218 -5.30495311\n",
+ " -5.44473025 -5.35376888 -5.39470442 -5.39844245 -5.5415277 -5.62739757\n",
+ " -5.55218576 -5.52254264 -5.56258082 -5.48827193 -5.57307772 -5.64981534\n",
+ " -5.64923964 -5.74919155 -5.55077531 -5.52218954 -5.83158053 -5.53706381\n",
+ " -5.61278332 -5.55331891 -5.53050426 -5.66872213 -5.49724804 -5.51510606\n",
+ " -6.53225066 -6.64830432 -6.45494563 -6.48848104 -6.34773477 -6.6166628\n",
+ " -6.48395451 -6.37164209 -6.40806271 -6.4014871 -6.38524937 -6.42101867\n",
+ " -6.61623251 -6.47698936 -6.5687579 -6.53685093 -6.4563278 -6.49000115\n",
+ " -6.52431737 -6.5734751 -7.43909839 -7.26374908 -7.12908348 -6.99310143\n",
+ " -7.02973246 -6.8177417 -7.1870296 -7.06206145 -7.43004635 -7.04339076\n",
+ " -7.10411536 -7.21828719 -7.13995389 -7.15239091 -7.08778731 -7.03879194\n",
+ " -7.16154811 -6.96074688 -7.01796131 -7.03080328 nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
@@ -705,20 +732,20 @@
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
" nan nan nan nan nan nan\n",
- " -4.98448407 -5.02507457 -4.93247179 -4.92906139 -4.96748311 -4.94203403\n",
- " -4.9387752 -5.00661944 -4.94224026 -5.02843045 -4.92155128 -4.99690417\n",
- " -4.98389323 -4.96017288 -4.99919508 -4.93480677 -5.00840005 -5.01310808\n",
- " -4.95533999 -4.92532856 -5.41911845 -5.46486328 -5.54655607 -5.41261174\n",
- " -5.48783365 -5.30235912 -5.49721981 -5.50506168 -5.4623197 -5.50178558\n",
- " -5.4530093 -5.38804699 -5.38547187 -5.49229764 -5.32674485 -5.38296403\n",
- " -5.37389633 -5.45172146 -5.43905269 -5.4069051 -6.38556131 -6.413713\n",
- " -6.5692907 -6.3776548 -6.52825337 -6.22175329 -6.4196389 -6.39976861\n",
- " -6.55018874 -6.47026458 -6.53503514 -6.53618594 -6.26546457 -6.47244915\n",
- " -6.39532854 -6.40273941 -6.20885683 -6.42818791 -6.45390754 -6.43771307\n",
- " -6.6618988 -6.77896678 -6.6357699 -6.68441837 -6.65275642 -6.59599737\n",
- " -6.60110611 -6.70986504 -6.5456143 -6.89273932 -6.57695001 -6.59719718\n",
- " -6.6640911 -6.44351792 -6.60824781 -6.64304958 -6.62943501 -6.52970465\n",
- " -6.71383902 -6.67454936]\n",
+ " -5.08367846 -5.18535892 -5.15306679 -5.0897679 -5.24227234 -5.22648725\n",
+ " -5.10455269 -5.15119939 -5.07737567 -5.18528714 -5.07791844 -5.23650054\n",
+ " -5.26045529 -5.11444114 -5.20007612 -5.1548395 -5.25596173 -5.20630333\n",
+ " -5.15792876 -5.26337324 -5.60353316 -5.3637392 -5.40574354 -5.40940198\n",
+ " -5.44445189 -5.46083488 -5.41928986 -5.41994037 -5.49137874 -5.41926369\n",
+ " -5.45424213 -5.4073457 -5.37592626 -5.41369061 -5.4826103 -5.4304887\n",
+ " -5.55409814 -5.61826834 -5.45056639 -5.43011355 -6.53732648 -6.48348268\n",
+ " -6.41855927 -6.51234273 -6.42253816 -6.43424793 -6.46200622 -6.47889909\n",
+ " -6.51187627 -6.41473923 -6.31999138 -6.52160562 -6.81179198 -6.48094732\n",
+ " -6.28119831 -6.31623776 -6.59911147 -6.30995227 -6.30666572 -6.43595374\n",
+ " -7.31528174 -6.81912515 -7.02646759 -7.08670358 -7.18273333 -7.10314959\n",
+ " -7.05651837 -7.05816259 -7.18931309 -7.13776984 -6.92767401 -6.99449498\n",
+ " -7.06979048 -7.11022562 -7.09744469 -7.02986576 -6.85218872 -7.24356397\n",
+ " -6.97618259 -7.07550066]\n",
" warnings.warn(\n"
]
}
@@ -781,7 +808,7 @@
"name": "stdout",
"output_type": "stream",
"text": [
- "GradientBoostingRegressor was finalized in 7.3396 seconds\n"
+ "GradientBoostingRegressor was finalized in 7.1047 seconds\n"
]
},
{
@@ -798,8 +825,8 @@
"name": "stdout",
"output_type": "stream",
"text": [
- "AdaBoostRegressor was finalized in 1.7686 seconds\n",
- "KNeighborsRegressor was finalized in 0.0357 seconds\n"
+ "AdaBoostRegressor was finalized in 1.8835 seconds\n",
+ "KNeighborsRegressor was finalized in 0.0332 seconds\n"
]
},
{
@@ -873,8 +900,6 @@
"/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/neural_network/_multilayer_perceptron.py:691: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (750) reached and the optimization hasn't converged yet.\n",
" warnings.warn(\n",
"/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/neural_network/_multilayer_perceptron.py:691: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (750) reached and the optimization hasn't converged yet.\n",
- " warnings.warn(\n",
- "/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/neural_network/_multilayer_perceptron.py:691: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (750) reached and the optimization hasn't converged yet.\n",
" warnings.warn(\n"
]
},
@@ -1025,6 +1050,8 @@
"/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/neural_network/_multilayer_perceptron.py:691: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (750) reached and the optimization hasn't converged yet.\n",
" warnings.warn(\n",
"/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/neural_network/_multilayer_perceptron.py:691: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (750) reached and the optimization hasn't converged yet.\n",
+ " warnings.warn(\n",
+ "/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/neural_network/_multilayer_perceptron.py:691: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (750) reached and the optimization hasn't converged yet.\n",
" warnings.warn(\n"
]
},
@@ -1099,6 +1126,8 @@
"/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/neural_network/_multilayer_perceptron.py:691: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (750) reached and the optimization hasn't converged yet.\n",
" warnings.warn(\n",
"/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/neural_network/_multilayer_perceptron.py:691: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (750) reached and the optimization hasn't converged yet.\n",
+ " warnings.warn(\n",
+ "/home/sosamuli/anaconda3/lib/python3.10/site-packages/sklearn/neural_network/_multilayer_perceptron.py:691: ConvergenceWarning: Stochastic Optimizer: Maximum iterations (750) reached and the optimization hasn't converged yet.\n",
" warnings.warn(\n"
]
},
@@ -1214,8 +1243,8 @@
"name": "stdout",
"output_type": "stream",
"text": [
- "MLPRegressor was finalized in 106.0161 seconds\n",
- "XGBRegressor was finalized in 5.5979 seconds\n"
+ "MLPRegressor was finalized in 67.2607 seconds\n",
+ "XGBRegressor was finalized in 4.6800 seconds\n"
]
}
],
@@ -1286,7 +1315,7 @@
},
{
"data": {
- "image/png": "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\n",
+ "image/png": "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\n",
"text/plain": [
"