workbench 0.8.178__py3-none-any.whl → 0.8.180__py3-none-any.whl

workbench/model_scripts/xgb_model/generated_model_script.py

@@ -32,10 +32,12 @@ TEMPLATE_PARAMS = {
     "target": "udm_asy_res_value",
     "features": ['chi2v', 'fr_sulfone', 'chi1v', 'bcut2d_logplow', 'fr_piperzine', 'kappa3', 'smr_vsa1', 'slogp_vsa5', 'fr_ketone_topliss', 'fr_sulfonamd', 'fr_imine', 'fr_benzene', 'fr_ester', 'chi2n', 'labuteasa', 'peoe_vsa2', 'smr_vsa6', 'bcut2d_chglo', 'fr_sh', 'peoe_vsa1', 'fr_allylic_oxid', 'chi4n', 'fr_ar_oh', 'fr_nh0', 'fr_term_acetylene', 'slogp_vsa7', 'slogp_vsa4', 'estate_vsa1', 'vsa_estate4', 'numbridgeheadatoms', 'numheterocycles', 'fr_ketone', 'fr_morpholine', 'fr_guanido', 'estate_vsa2', 'numheteroatoms', 'fr_nitro_arom_nonortho', 'fr_piperdine', 'nocount', 'numspiroatoms', 'fr_aniline', 'fr_thiophene', 'slogp_vsa10', 'fr_amide', 'slogp_vsa2', 'fr_epoxide', 'vsa_estate7', 'fr_ar_coo', 'fr_imidazole', 'fr_nitrile', 'fr_oxazole', 'numsaturatedrings', 'fr_pyridine', 'fr_hoccn', 'fr_ndealkylation1', 'numaliphaticheterocycles', 'fr_phenol', 'maxpartialcharge', 'vsa_estate5', 'peoe_vsa13', 'minpartialcharge', 'qed', 'fr_al_oh', 'slogp_vsa11', 'chi0n', 'fr_bicyclic', 'peoe_vsa12', 'fpdensitymorgan1', 'fr_oxime', 'molwt', 'fr_dihydropyridine', 'smr_vsa5', 'peoe_vsa5', 'fr_nitro', 'hallkieralpha', 'heavyatommolwt', 'fr_alkyl_halide', 'peoe_vsa8', 'fr_nhpyrrole', 'fr_isocyan', 'bcut2d_chghi', 'fr_lactam', 'peoe_vsa11', 'smr_vsa9', 'tpsa', 'chi4v', 'slogp_vsa1', 'phi', 'bcut2d_logphi', 'avgipc', 'estate_vsa11', 'fr_coo', 'bcut2d_mwhi', 'numunspecifiedatomstereocenters', 'vsa_estate10', 'estate_vsa8', 'numvalenceelectrons', 'fr_nh2', 'fr_lactone', 'vsa_estate1', 'estate_vsa4', 'numatomstereocenters', 'vsa_estate8', 'fr_para_hydroxylation', 'peoe_vsa3', 'fr_thiazole', 'peoe_vsa10', 'fr_ndealkylation2', 'slogp_vsa12', 'peoe_vsa9', 'maxestateindex', 'fr_quatn', 'smr_vsa7', 'minestateindex', 'numaromaticheterocycles', 'numrotatablebonds', 'fr_ar_nh', 'fr_ether', 'exactmolwt', 'fr_phenol_noorthohbond', 'slogp_vsa3', 'fr_ar_n', 'sps', 'fr_c_o_nocoo', 'bertzct', 'peoe_vsa7', 'slogp_vsa8', 'numradicalelectrons', 'molmr', 'fr_tetrazole', 'numsaturatedcarbocycles', 'bcut2d_mrhi', 'kappa1', 'numamidebonds', 'fpdensitymorgan2', 'smr_vsa8', 'chi1n', 'estate_vsa6', 'fr_barbitur', 'fr_diazo', 'kappa2', 'chi0', 'bcut2d_mrlow', 'balabanj', 'peoe_vsa4', 'numhacceptors', 'fr_sulfide', 'chi3n', 'smr_vsa2', 'fr_al_oh_notert', 'fr_benzodiazepine', 'fr_phos_ester', 'fr_aldehyde', 'fr_coo2', 'estate_vsa5', 'fr_prisulfonamd', 'numaromaticcarbocycles', 'fr_unbrch_alkane', 'fr_urea', 'fr_nitroso', 'smr_vsa10', 'fr_c_s', 'smr_vsa3', 'fr_methoxy', 'maxabspartialcharge', 'slogp_vsa9', 'heavyatomcount', 'fr_azide', 'chi3v', 'smr_vsa4', 'mollogp', 'chi0v', 'fr_aryl_methyl', 'fr_nh1', 'fpdensitymorgan3', 'fr_furan', 'fr_hdrzine', 'fr_arn', 'numaromaticrings', 'vsa_estate3', 'fr_azo', 'fr_halogen', 'estate_vsa9', 'fr_hdrzone', 'numhdonors', 'fr_alkyl_carbamate', 'fr_isothiocyan', 'minabspartialcharge', 'fr_al_coo', 'ringcount', 'chi1', 'estate_vsa7', 'fr_nitro_arom', 'vsa_estate9', 'minabsestateindex', 'maxabsestateindex', 'vsa_estate6', 'estate_vsa10', 'estate_vsa3', 'fr_n_o', 'fr_amidine', 'fr_thiocyan', 'fr_phos_acid', 'fr_c_o', 'fr_imide', 'numaliphaticrings', 'peoe_vsa6', 'vsa_estate2', 'nhohcount', 'numsaturatedheterocycles', 'slogp_vsa6', 'peoe_vsa14', 'fractioncsp3', 'bcut2d_mwlow', 'numaliphaticcarbocycles', 'fr_priamide', 'nacid', 'nbase', 'naromatom', 'narombond', 'sz', 'sm', 'sv', 'sse', 'spe', 'sare', 'sp', 'si', 'mz', 'mm', 'mv', 'mse', 'mpe', 'mare', 'mp', 'mi', 'xch_3d', 'xch_4d', 'xch_5d', 'xch_6d', 'xch_7d', 'xch_3dv', 'xch_4dv', 'xch_5dv', 'xch_6dv', 'xch_7dv', 'xc_3d', 'xc_4d', 'xc_5d', 'xc_6d', 'xc_3dv', 'xc_4dv', 'xc_5dv', 'xc_6dv', 'xpc_4d', 'xpc_5d', 'xpc_6d', 'xpc_4dv', 'xpc_5dv', 'xpc_6dv', 'xp_0d', 'xp_1d', 'xp_2d', 'xp_3d', 'xp_4d', 'xp_5d', 'xp_6d', 'xp_7d', 'axp_0d', 'axp_1d', 'axp_2d', 'axp_3d', 'axp_4d', 'axp_5d', 'axp_6d', 'axp_7d', 'xp_0dv', 'xp_1dv', 'xp_2dv', 'xp_3dv', 'xp_4dv', 'xp_5dv', 'xp_6dv', 'xp_7dv', 'axp_0dv', 'axp_1dv', 'axp_2dv', 'axp_3dv', 'axp_4dv', 'axp_5dv', 'axp_6dv', 'axp_7dv', 'c1sp1', 'c2sp1', 'c1sp2', 'c2sp2', 'c3sp2', 'c1sp3', 'c2sp3', 'c3sp3', 'c4sp3', 'hybratio', 'fcsp3', 'num_stereocenters', 'num_unspecified_stereocenters', 'num_defined_stereocenters', 'num_r_centers', 'num_s_centers', 'num_stereobonds', 'num_e_bonds', 'num_z_bonds', 'stereo_complexity', 'frac_defined_stereo'],
     "compressed_features": [],
-    "model_metrics_s3_path": "s3://ideaya-sageworks-bucket/models/pka-a1-reg-0-nightly-100-test/training",
-    "train_all_data": True
+    "model_metrics_s3_path": "s3://ideaya-sageworks-bucket/models/logd-hyper-80/training",
+    "train_all_data": False,
+    "hyperparameters": {'n_estimators': 200, 'max_depth': 6, 'learning_rate': 0.05, 'subsample': 0.7, 'colsample_bytree': 0.3, 'colsample_bylevel': 0.5, 'min_child_weight': 5, 'gamma': 0.2, 'reg_alpha': 0.5, 'reg_lambda': 2.0, 'scale_pos_weight': 1},
 }
 
+
 # Function to check if dataframe is empty
 def check_dataframe(df: pd.DataFrame, df_name: str) -> None:
     """
@@ -75,7 +77,7 @@ def expand_proba_column(df: pd.DataFrame, class_labels: List[str]) -> pd.DataFra
     proba_df = pd.DataFrame(df[proba_column].tolist(), columns=proba_splits)
 
     # Drop any proba columns and reset the index in prep for the concat
-    df = df.drop(columns=[proba_column]+proba_splits, errors="ignore")
+    df = df.drop(columns=[proba_column] + proba_splits, errors="ignore")
     df = df.reset_index(drop=True)
 
     # Concatenate the new columns with the original DataFrame
@@ -140,8 +142,10 @@ def convert_categorical_types(df: pd.DataFrame, features: list, category_mapping
     return df, category_mappings
 
 
-def decompress_features(df: pd.DataFrame, features: List[str], compressed_features: List[str]) -> Tuple[pd.DataFrame, List[str]]:
-    """Prepare features for the XGBoost model
+def decompress_features(
+    df: pd.DataFrame, features: List[str], compressed_features: List[str]
+) -> Tuple[pd.DataFrame, List[str]]:
+    """Prepare features for the model
 
     Args:
         df (pd.DataFrame): The features DataFrame
@@ -204,6 +208,7 @@ if __name__ == "__main__":
     model_type = TEMPLATE_PARAMS["model_type"]
     model_metrics_s3_path = TEMPLATE_PARAMS["model_metrics_s3_path"]
     train_all_data = TEMPLATE_PARAMS["train_all_data"]
+    hyperparameters = TEMPLATE_PARAMS["hyperparameters"]
     validation_split = 0.2
 
     # Script arguments for input/output directories
@@ -216,11 +221,7 @@ if __name__ == "__main__":
     args = parser.parse_args()
 
     # Read the training data into DataFrames
-    training_files = [
-        os.path.join(args.train, file)
-        for file in os.listdir(args.train)
-        if file.endswith(".csv")
-    ]
+    training_files = [os.path.join(args.train, file) for file in os.listdir(args.train) if file.endswith(".csv")]
     print(f"Training Files: {training_files}")
 
     # Combine files and read them all into a single pandas dataframe
@@ -255,15 +256,16 @@ if __name__ == "__main__":
     else:
         # Just do a random training Split
         print("WARNING: No training column found, splitting data with random state=42")
-        df_train, df_val = train_test_split(
-            all_df, test_size=validation_split, random_state=42
-        )
+        df_train, df_val = train_test_split(all_df, test_size=validation_split, random_state=42)
     print(f"FIT/TRAIN: {df_train.shape}")
     print(f"VALIDATION: {df_val.shape}")
 
+    # Use any hyperparameters to set up both the trainer and model configurations
+    print(f"Hyperparameters: {hyperparameters}")
+
     # Now spin up our XGB Model
     if model_type == "classifier":
-        xgb_model = xgb.XGBClassifier(enable_categorical=True)
+        xgb_model = xgb.XGBClassifier(enable_categorical=True, **hyperparameters)
 
         # Encode the target column
         label_encoder = LabelEncoder()
@@ -271,12 +273,12 @@ if __name__ == "__main__":
         df_val[target] = label_encoder.transform(df_val[target])
 
     else:
-        xgb_model = xgb.XGBRegressor(enable_categorical=True)
+        xgb_model = xgb.XGBRegressor(enable_categorical=True, **hyperparameters)
         label_encoder = None  # We don't need this for regression
 
     # Grab our Features, Target and Train the Model
     y_train = df_train[target]
-    X_train= df_train[features]
+    X_train = df_train[features]
     xgb_model.fit(X_train, y_train)
 
     # Make Predictions on the Validation Set
@@ -315,9 +317,7 @@ if __name__ == "__main__":
         label_names = label_encoder.classes_
 
         # Calculate various model performance metrics
-        scores = precision_recall_fscore_support(
-            y_validate, preds, average=None, labels=label_names
-        )
+        scores = precision_recall_fscore_support(y_validate, preds, average=None, labels=label_names)
 
         # Put the scores into a dataframe
         score_df = pd.DataFrame(
@@ -355,7 +355,9 @@ if __name__ == "__main__":
         print(f"NumRows: {len(df_val)}")
 
     # Now save the model to the standard place/name
-    xgb_model.save_model(os.path.join(args.model_dir, "xgb_model.json"))
+    joblib.dump(xgb_model, os.path.join(args.model_dir, "xgb_model.joblib"))
+
+    # Save the label encoder if we have one
     if label_encoder:
         joblib.dump(label_encoder, os.path.join(args.model_dir, "label_encoder.joblib"))
 
@@ -370,19 +372,8 @@ if __name__ == "__main__":
 
 def model_fn(model_dir):
     """Deserialize and return fitted XGBoost model"""
-
-    model_path = os.path.join(model_dir, "xgb_model.json")
-
-    with open(model_path, "r") as f:
-        model_json = json.load(f)
-
-    sklearn_data = model_json['learner']['attributes']['scikit_learn']
-    model_type = json.loads(sklearn_data)['_estimator_type']
-
-    model_class = xgb.XGBClassifier if model_type == "classifier" else xgb.XGBRegressor
-    model = model_class(enable_categorical=True)
-    model.load_model(model_path)
-
+    model_path = os.path.join(model_dir, "xgb_model.joblib")
+    model = joblib.load(model_path)
     return model
 
 

workbench/model_scripts/xgb_model/xgb_model.template

@@ -33,9 +33,11 @@ TEMPLATE_PARAMS = {
     "features": "{{feature_list}}",
     "compressed_features": "{{compressed_features}}",
     "model_metrics_s3_path": "{{model_metrics_s3_path}}",
-    "train_all_data": "{{train_all_data}}"
+    "train_all_data": "{{train_all_data}}",
+    "hyperparameters": "{{hyperparameters}}",
 }
 
+
 # Function to check if dataframe is empty
 def check_dataframe(df: pd.DataFrame, df_name: str) -> None:
     """
@@ -75,7 +77,7 @@ def expand_proba_column(df: pd.DataFrame, class_labels: List[str]) -> pd.DataFra
     proba_df = pd.DataFrame(df[proba_column].tolist(), columns=proba_splits)
 
     # Drop any proba columns and reset the index in prep for the concat
-    df = df.drop(columns=[proba_column]+proba_splits, errors="ignore")
+    df = df.drop(columns=[proba_column] + proba_splits, errors="ignore")
     df = df.reset_index(drop=True)
 
     # Concatenate the new columns with the original DataFrame
@@ -140,8 +142,10 @@ def convert_categorical_types(df: pd.DataFrame, features: list, category_mapping
     return df, category_mappings
 
 
-def decompress_features(df: pd.DataFrame, features: List[str], compressed_features: List[str]) -> Tuple[pd.DataFrame, List[str]]:
-    """Prepare features for the XGBoost model
+def decompress_features(
+    df: pd.DataFrame, features: List[str], compressed_features: List[str]
+) -> Tuple[pd.DataFrame, List[str]]:
+    """Prepare features for the model
 
     Args:
         df (pd.DataFrame): The features DataFrame
@@ -204,6 +208,7 @@ if __name__ == "__main__":
     model_type = TEMPLATE_PARAMS["model_type"]
     model_metrics_s3_path = TEMPLATE_PARAMS["model_metrics_s3_path"]
     train_all_data = TEMPLATE_PARAMS["train_all_data"]
+    hyperparameters = TEMPLATE_PARAMS["hyperparameters"]
     validation_split = 0.2
 
     # Script arguments for input/output directories
@@ -216,11 +221,7 @@ if __name__ == "__main__":
     args = parser.parse_args()
 
     # Read the training data into DataFrames
-    training_files = [
-        os.path.join(args.train, file)
-        for file in os.listdir(args.train)
-        if file.endswith(".csv")
-    ]
+    training_files = [os.path.join(args.train, file) for file in os.listdir(args.train) if file.endswith(".csv")]
     print(f"Training Files: {training_files}")
 
     # Combine files and read them all into a single pandas dataframe
@@ -255,15 +256,16 @@ if __name__ == "__main__":
     else:
         # Just do a random training Split
         print("WARNING: No training column found, splitting data with random state=42")
-        df_train, df_val = train_test_split(
-            all_df, test_size=validation_split, random_state=42
-        )
+        df_train, df_val = train_test_split(all_df, test_size=validation_split, random_state=42)
     print(f"FIT/TRAIN: {df_train.shape}")
     print(f"VALIDATION: {df_val.shape}")
 
+    # Use any hyperparameters to set up both the trainer and model configurations
+    print(f"Hyperparameters: {hyperparameters}")
+
     # Now spin up our XGB Model
     if model_type == "classifier":
-        xgb_model = xgb.XGBClassifier(enable_categorical=True)
+        xgb_model = xgb.XGBClassifier(enable_categorical=True, **hyperparameters)
 
         # Encode the target column
         label_encoder = LabelEncoder()
@@ -271,12 +273,12 @@ if __name__ == "__main__":
         df_val[target] = label_encoder.transform(df_val[target])
 
     else:
-        xgb_model = xgb.XGBRegressor(enable_categorical=True)
+        xgb_model = xgb.XGBRegressor(enable_categorical=True, **hyperparameters)
         label_encoder = None  # We don't need this for regression
 
     # Grab our Features, Target and Train the Model
     y_train = df_train[target]
-    X_train= df_train[features]
+    X_train = df_train[features]
     xgb_model.fit(X_train, y_train)
 
     # Make Predictions on the Validation Set
@@ -315,9 +317,7 @@ if __name__ == "__main__":
         label_names = label_encoder.classes_
 
         # Calculate various model performance metrics
-        scores = precision_recall_fscore_support(
-            y_validate, preds, average=None, labels=label_names
-        )
+        scores = precision_recall_fscore_support(y_validate, preds, average=None, labels=label_names)
 
         # Put the scores into a dataframe
         score_df = pd.DataFrame(
@@ -355,7 +355,9 @@ if __name__ == "__main__":
         print(f"NumRows: {len(df_val)}")
 
     # Now save the model to the standard place/name
-    xgb_model.save_model(os.path.join(args.model_dir, "xgb_model.json"))
+    joblib.dump(xgb_model, os.path.join(args.model_dir, "xgb_model.joblib"))
+
+    # Save the label encoder if we have one
     if label_encoder:
         joblib.dump(label_encoder, os.path.join(args.model_dir, "label_encoder.joblib"))
 
@@ -370,19 +372,8 @@ if __name__ == "__main__":
 
 def model_fn(model_dir):
     """Deserialize and return fitted XGBoost model"""
-
-    model_path = os.path.join(model_dir, "xgb_model.json")
-
-    with open(model_path, "r") as f:
-        model_json = json.load(f)
-
-    sklearn_data = model_json['learner']['attributes']['scikit_learn']
-    model_type = json.loads(sklearn_data)['_estimator_type']
-
-    model_class = xgb.XGBClassifier if model_type == "classifier" else xgb.XGBRegressor
-    model = model_class(enable_categorical=True)
-    model.load_model(model_path)
-
+    model_path = os.path.join(model_dir, "xgb_model.joblib")
+    model = joblib.load(model_path)
     return model
 
 

workbench/utils/model_utils.py

@@ -222,7 +222,8 @@ def uq_metrics(df: pd.DataFrame, target_col: str) -> Dict[str, Any]:
         lower_95, upper_95 = df["q_025"], df["q_975"]
         lower_90, upper_90 = df["q_05"], df["q_95"]
         lower_80, upper_80 = df["q_10"], df["q_90"]
-        lower_68, upper_68 = df["q_16"], df["q_84"]
+        lower_68 = df.get("q_16", 0)
+        upper_68 = df.get("q_84", 0)
         lower_50, upper_50 = df["q_25"], df["q_75"]
     elif "prediction_std" in df.columns:
         lower_95 = df["prediction"] - 1.96 * df["prediction_std"]

workbench/utils/shap_utils.py

@@ -212,6 +212,14 @@ def _calculate_shap_values(workbench_model, sample_df: pd.DataFrame = None):
         log.error("No XGBoost model found in the artifact.")
         return None, None, None, None
 
+    # Get the booster (SHAP requires the booster, not the sklearn wrapper)
+    if hasattr(xgb_model, "get_booster"):
+        # Full sklearn model - extract the booster
+        booster = xgb_model.get_booster()
+    else:
+        # Already a booster
+        booster = xgb_model
+
     # Load category mappings if available
     category_mappings = load_category_mappings_from_s3(model_artifact_uri)
 
@@ -229,8 +237,8 @@ def _calculate_shap_values(workbench_model, sample_df: pd.DataFrame = None):
     # Create a DMatrix with categorical support
     dmatrix = xgb.DMatrix(X, enable_categorical=True)
 
-    # Use XGBoost's built-in SHAP calculation
-    shap_values = xgb_model.predict(dmatrix, pred_contribs=True, strict_shape=True)
+    # Use XGBoost's built-in SHAP calculation (booster method, not sklearn)
+    shap_values = booster.predict(dmatrix, pred_contribs=True, strict_shape=True)
     features_with_bias = features + ["bias"]
 
     # Now we need to subset the columns based on top 10 SHAP values