update changes

Author: zyliang2001

feature_importance/01_ablation_regression_script.sh

@@ -5,7 +5,7 @@
 
 source activate mdi
 # Need to specify --result_name --ablate_features(default all features) --fitted(default not fitted)
-command="01_run_ablation_regression.py --nreps 1 --config mdi_local.real_data_regression --split_seed ${1} --ignore_cache --create_rmd --folder_name diabetes_final --fit_model True --positive_masking True --absolute_masking True --negative_masking True"
+command="01_run_ablation_regression.py --nreps 1 --config mdi_local.real_data_regression --split_seed ${1} --ignore_cache --create_rmd --folder_name diabetes_new_methods --fit_model True --positive_masking True --absolute_masking True --negative_masking True"
 
 # Execute the command
 python $command

feature_importance/01_ablation_script_regr.sh

@@ -1,6 +1,6 @@
 #!/bin/bash
 
-slurm_script="01_ablation_regression_script_synthetic.sh"
+slurm_script="01_ablation_regression_script.sh"
 
 for rep in {1..10}
 do

feature_importance/01_auroc_regression_script_linear.sh

@@ -5,7 +5,7 @@
 
 source activate mdi
 # Need to specify --result_name --ablate_features(default all features) --fitted(default not fitted)
-command="01_run_auroc_synthetic.py --nreps 1 --config mdi_local.synthetic_data_linear --split_seed 0 --simulation_seed ${1} --ignore_cache --create_rmd --folder_name linear_synthetic_new --fit_model True"
+command="01_run_auroc_synthetic.py --nreps 1 --config mdi_local.synthetic_data_linear --split_seed 0 --simulation_seed ${1} --ignore_cache --create_rmd --folder_name linear_one_group_test_300_avg_leaf --fit_model True"
 
 # Execute the command
 python $command

feature_importance/01_auroc_regression_script_linear2.sh

@@ -0,0 +1,11 @@
+#!/bin/bash
+#SBATCH --mail-user=zhongyuan_liang@berkeley.edu
+#SBATCH --mail-type=ALL
+#SBATCH --partition=yugroup
+
+source activate mdi
+# Need to specify --result_name --ablate_features(default all features) --fitted(default not fitted)
+command="01_run_auroc_synthetic.py --nreps 1 --config mdi_local.synthetic_data_linear2 --split_seed 0 --simulation_seed ${1} --ignore_cache --create_rmd --folder_name linear_one_group_2 --fit_model True"
+
+# Execute the command
+python $command

feature_importance/01_auroc_regression_script_linear_concept_shift.sh

@@ -0,0 +1,11 @@
+#!/bin/bash
+#SBATCH --mail-user=zhongyuan_liang@berkeley.edu
+#SBATCH --mail-type=ALL
+#SBATCH --partition=yugroup
+
+source activate mdi
+# Need to specify --result_name --ablate_features(default all features) --fitted(default not fitted)
+command="01_run_auroc_synthetic.py --nreps 1 --config mdi_local.synthetic_data_linear_concept_shift --split_seed 0 --simulation_seed ${1} --ignore_cache --create_rmd --folder_name linear_two_groups_concept_shift_test_300 --fit_model True"
+
+# Execute the command
+python $command

feature_importance/01_auroc_regression_script_linear_concept_shift2.sh

@@ -0,0 +1,11 @@
+#!/bin/bash
+#SBATCH --mail-user=zhongyuan_liang@berkeley.edu
+#SBATCH --mail-type=ALL
+#SBATCH --partition=yugroup
+
+source activate mdi
+# Need to specify --result_name --ablate_features(default all features) --fitted(default not fitted)
+command="01_run_auroc_synthetic.py --nreps 1 --config mdi_local.synthetic_data_linear_concept_shift2 --split_seed 0 --simulation_seed ${1} --ignore_cache --create_rmd --folder_name linear_two_groups_concept_shift_2 --fit_model True"
+
+# Execute the command
+python $command

feature_importance/01_auroc_regression_script_lss.sh

@@ -5,7 +5,7 @@
 
 source activate mdi
 # Need to specify --result_name --ablate_features(default all features) --fitted(default not fitted)
-command="01_run_auroc_synthetic.py --nreps 1 --config mdi_local.synthetic_data_lss --split_seed ${1} --ignore_cache --create_rmd --folder_name lss_synthetic --fit_model True"
+command="01_run_auroc_synthetic_lss.py --nreps 1 --config mdi_local.synthetic_data_lss --split_seed 0 --simulation_seed ${1} --ignore_cache --create_rmd --folder_name lss_one_group_test_300_avg_leaf --fit_model True"
 
 # Execute the command
 python $command

feature_importance/01_auroc_script_regr.sh

@@ -1,8 +1,8 @@
 #!/bin/bash
 
-slurm_script="01_auroc_regression_script_linear.sh"
+slurm_script="01_auroc_regression_script_lss.sh" #"01_auroc_regression_script_linear_concept_shift.sh"
 
-for rep in {1..3}
+for rep in {1..10}
 do
     sbatch $slurm_script $rep  # Submit SLURM job using the specified script
     sleep 2

feature_importance/01_run_ablation_regression.py

@@ -289,14 +289,19 @@ def compare_estimators(estimators: List[ModelConfig],
                         mode.append("positive")
                     if args.negative_masking:
                         mode.append("negative")
+
+                    if loaded_model is not None:
+                        y_train_pred = loaded_model.predict(X_train)
+                    else:
+                        y_train_pred = None
 
                     for m in mode:
                         start = time.time()
                         print(f"Compute feature importance")
                         # Compute feature importance
                         local_fi_score_train, local_fi_score_train_subset, local_fi_score_test, local_fi_score_test_subset = fi_est.cls(X_train=X_train, y_train=y_train, X_train_subset = X_train_subset, y_train_subset=y_train_subset,
                                                                                                                                         X_test=X_test, y_test=y_test, X_test_subset=X_test_subset, y_test_subset=y_test_subset,
-                                                                                                                                        fit=loaded_model, mode=m)
+                                                                                                                                        fit=loaded_model, mode=m, y_train_pred=y_train_pred)
                         if fi_est.name.startswith("Local_MDI+"):
                             local_fi_score_train_subset = local_fi_score_train[indices_train]