teradataml 20.0.0.1__py3-none-any.whl → 20.0.0.2__py3-none-any.whl

teradataml/automl/model_evaluation.py

@@ -15,6 +15,7 @@
 
 # Python libraries
 import time
+import ast
 
 # Teradata libraries
 from teradataml.dataframe.dataframe import DataFrame
@@ -56,7 +57,12 @@ class _ModelEvaluator:
         self.target_column = target_column
         self.task_type = task_type
         
-    def model_evaluation(self, rank, table_name_mapping, test_data_ind = False, target_column_ind = False):
+    def model_evaluation(self, 
+                         rank, 
+                         table_name_mapping,
+                         data_node_id, 
+                         target_column_ind = True,
+                         get_metrics = False):
         """
         DESCRIPTION:
             Function performs the model evaluation on the specified rank in leaderborad.
@@ -72,25 +78,32 @@ class _ModelEvaluator:
                 Specifies the mapping of train,test table names.
                 Types: dict
             
-            test_data_ind:
-                Optional Argument.
-                Specifies whether test data is present or not.
-                Default Value: False
-                Types: bool
+            data_node_id:
+                Required Argument.
+                Specifies the test data node id.
+                Types: str
             
             target_column_ind:
                 Optional Argument.
                 Specifies whether target column is present in the dataset or not.
+                Default Value: True
+                Types: bool
+      
+            get_metrics:
+                Optional Argument.
+                Specifies whether to return metrics or not.
                 Default Value: False
+                Types: bool
                 
         RETURNS:
             tuple containing, performance metrics and predicitions of specified rank ML model.
              
         """
-        # Setting test data indicator and target column indicator
-        self.test_data_ind = test_data_ind
+        # Setting target column indicator
         self.target_column_ind = target_column_ind
         self.table_name_mapping = table_name_mapping
+        self.data_node_id = data_node_id
+        self.get_metrics = get_metrics
         
         # Return predictions only if test data is present and target column is not present
         return self._evaluator(rank)
@@ -114,38 +127,34 @@ class _ModelEvaluator:
         """
         # Extracting model using rank
         model = self.model_info.loc[rank]
+
+        ml_name = self.model_info.loc[rank]['MODEL_ID'].split('_')[0]
         
         # Defining eval_params 
-        eval_params = _ModelTraining._eval_params_generation(model['Name'],
+        eval_params = _ModelTraining._eval_params_generation(ml_name,
                                                              self.target_column,
                                                              self.task_type)
         
-        # Test Data
-        test = DataFrame(self.table_name_mapping['{}_test'.format(model['Feature-Selection'])])
-
-        # Getting test data from table
-        if not self.test_data_ind:
-            # Test Data
-            test = DataFrame(self.table_name_mapping['{}_test'.format(model['Feature-Selection'])])
-        else:
-            test = DataFrame(self.table_name_mapping['{}_new_test'.format(model['Feature-Selection'])])
-
-        print("\nFollowing model is being used for generating prediction :")
-        print("Model ID :", model['Model-ID'],
-              "\nFeature Selection Method :",model['Feature-Selection'])
+        # Extracting test data for evaluation based on data node id
+        test = DataFrame(self.table_name_mapping[self.data_node_id]['{}_new_test'.format(model['FEATURE_SELECTION'])])
 
-        # Evaluation and predictions 
-        if model['Name'] == 'knn':
-            metrics = model['model-obj'].evaluate(test_data=test)
-            pred = model['model-obj'].predict(test_data=test)
+        print("\nFollowing model is being picked for evaluation:")
+        print("Model ID :", model['MODEL_ID'],
+              "\nFeature Selection Method :",model['FEATURE_SELECTION'])
+        
+        if self.task_type.lower() == 'classification':
+            params = ast.literal_eval(model['PARAMETERS'])
+            eval_params['output_responses'] = params['output_responses']
+        
+        # Mapping data according to model type
+        data_map = 'test_data' if ml_name == 'KNN' else 'newdata'
+        # Performing evaluation if get_metrics is True else returning predictions
+        if self.get_metrics:
+            metrics = model['model-obj'].evaluate(**{data_map: test}, **eval_params)
+            return metrics
         else:
-            # Return predictions only if test data is present and target column is not present
-            if self.test_data_ind and not self.target_column_ind:
+            # Removing accumulate parameter if target column is not present
+            if not self.target_column_ind:
                 eval_params.pop("accumulate")
-                pred = model['model-obj'].predict(newdata=test, **eval_params)
-                return pred
-            # Return both metrics and predictions for all other cases
-            metrics = model['model-obj'].evaluate(newdata=test, **eval_params)
-            pred = model['model-obj'].predict(newdata=test, **eval_params)
-        
-        return (metrics, pred)
+            pred = model['model-obj'].predict(**{data_map: test}, **eval_params)
+            return pred

teradataml/automl/model_training.py

@@ -19,6 +19,7 @@ from concurrent.futures import ThreadPoolExecutor
 import math
 import pandas as pd
 from itertools import product
+import numpy as np
 
 # Teradata libraries
 from teradataml.context import context as tdmlctx
@@ -38,7 +39,8 @@ class _ModelTraining:
                  verbose=0,
                  features=None,
                  task_type="Regression",
-                 custom_data = None):
+                 custom_data = None,
+                 **kwargs):
         """
         DESCRIPTION:
             Function initializes the data, target column, features and models
@@ -89,6 +91,28 @@ class _ModelTraining:
                 Optional Argument.
                 Specifies json object containing user customized input.
                 Types: json object
+
+            **kwargs:
+                Specifies the additional arguments for model training. Below
+                are the additional arguments:
+                    volatile:
+                        Optional Argument.
+                        Specifies whether to put the interim results of the
+                        functions in a volatile table or not. When set to
+                        True, results are stored in a volatile table,
+                        otherwise not.
+                        Default Value: False
+                        Types: bool
+
+                    persist:
+                        Optional Argument.
+                        Specifies whether to persist the interim results of the
+                        functions in a table or not. When set to True,
+                        results are persisted in a table; otherwise,
+                        results are garbage collected at the end of the
+                        session.
+                        Default Value: False
+                        Types: bool
         """
         self.data = data
         self.target_column = target_column
@@ -99,6 +123,8 @@ class _ModelTraining:
         self.custom_data = custom_data
         self.labels = self.data.drop_duplicate(self.target_column).size
         self.startify_col = None
+        self.persist = kwargs.get("persist", False)
+        self.volatile = kwargs.get("volatile", False)
 
     def model_training(self, 
                        auto=True,
@@ -278,20 +304,25 @@ class _ModelTraining:
         """	
         # Creating a copy to avoid use of same reference of memory
         if self.task_type != "Regression":
-            sorted_model_df = trained_models_info.sort_values(by=['Micro-F1', 'Weighted-F1'], 
-                                                ascending=[False, False]).reset_index(drop=True)
+            sorted_model_df = trained_models_info.sort_values(by=['MICRO-F1', 'WEIGHTED-F1'], 
+                                                              ascending=[False, False]).reset_index(drop=True)
         else:
-            sorted_model_df = trained_models_info.sort_values(by='R2-score', 
-                                                ascending=False).reset_index(drop=True)
+            sorted_model_df = trained_models_info.sort_values(by='R2', 
+                                                              ascending=False).reset_index(drop=True)
+
         
         # Adding rank to leaderboard
-        sorted_model_df.insert(0, 'Rank', sorted_model_df.index + 1) 
+        sorted_model_df.insert(0, 'RANK', sorted_model_df.index + 1) 
+
+        # Internal Data list for leaderboard
+        dp_lst = ["model-obj", "DATA_TABLE", "RESULT_TABLE", "PARAMETERS"]
 
         # Excluding the model object and model name from leaderboard
-        leaderboard = sorted_model_df.drop(["model-obj","Name"], axis=1)
+        leaderboard = sorted_model_df.drop(dp_lst, axis=1)
+
         # filtering the rows based on the max_models
         if self.max_models is not None:
-            leaderboard = leaderboard[leaderboard["Rank"] <= self.max_models]
+            leaderboard = leaderboard[leaderboard["RANK"] <= self.max_models]
         
         self._display_msg(msg="Leaderboard",
                           progress_bar=self.progress_bar,
@@ -436,28 +467,24 @@ class _ModelTraining:
             max_depth.extend([6, 7, 8])
             min_node_size.extend([2])
             iter_num.extend([20])
-            num_trees.extend([10, 20])
         elif num_rows < 10000 and num_cols < 15:
             min_impurity.extend([0.1, 0.2])
             shrinkage_factor.extend([0.1, 0.3])
             max_depth.extend([6, 8, 10])
             min_node_size.extend([2, 3])
             iter_num.extend([20, 30])
-            num_trees.extend([20, 30])
         elif num_rows < 100000 and num_cols < 20:
             min_impurity.extend([0.2, 0.3])
             shrinkage_factor.extend([0.01, 0.1, 0.2])
             max_depth.extend([4, 6, 7])
             min_node_size.extend([3, 4])
             iter_num.extend([30, 40])
-            num_trees.extend([30, 40])
         else:
             min_impurity.extend([0.1, 0.2, 0.3])
             shrinkage_factor.extend([0.01, 0.05, 0.1])
             max_depth.extend([3, 4, 7, 8])
             min_node_size.extend([2, 3, 4])
             iter_num.extend([20, 30, 40])
-            num_trees.extend([20, 30, 40])
 
         # Hyperparameters for XGBoost model
         xgb_params = {
@@ -736,12 +763,15 @@ class _ModelTraining:
 
         # Hyperparameters for each model
         model_params = parameters[:min(len(parameters), 5)]
-        self._display_msg(msg="\nPerforming hyperParameter tuning ...", progress_bar=self.progress_bar)
+        self._display_msg(msg="\nPerforming hyperparameter tuning ...", progress_bar=self.progress_bar)
 
-        # Defining training and testing data
+        # Defining training data
         data_types = ['lasso', 'rfe', 'pca']
         trainng_datas = tuple(DataFrame(self.table_name_mapping[f'{data_type}_train']) for data_type in data_types)
-        testing_datas = tuple(DataFrame(self.table_name_mapping[f'{data_type}_test']) for data_type in data_types)
+
+        if self.task_type == "Classification":
+            response_values = trainng_datas[0].get(self.target_column).drop_duplicate().get_values().flatten().tolist()
+            self.output_response = [str(i) for i in response_values]
 
         if self.stopping_metric is None:
             self.stopping_tolerance, self.stopping_metric = 1.0, 'MICRO-F1' \
@@ -765,115 +795,15 @@ class _ModelTraining:
 
         trained_models = []
         for param in model_params:
-            result = self._hyperparameter_tunning(param, trainng_datas, testing_datas)
+            result = self._hyperparameter_tunning(param, trainng_datas)
             trained_models.append(result)
 
         models_df = pd.concat(trained_models, ignore_index=True)
-
-        # Score the model and combine the results into a single DataFrame
-        trained_models_info = self._model_scoring(testing_datas, models_df)
-        trained_models_info = trained_models_info.reset_index(drop=True)
-
-        return trained_models_info
-    
-    def _model_scoring(self,
-                       test_data,
-                       model_info):
-        """
-        DESCRIPTION:
-            Internal function generates the performance metrics for
-            trained ML models using testing dataset.
-         
-        PARAMETERS:
-            test_data
-                Required Argument.
-                Specifies the testing datasets
-                Types: tuple of Teradataml DataFrame
-            
-            model_info
-                Required Argument.
-                Specifies the trained models information.
-                Types: Pandas DataFrame
-            
-        RETURNS:
-            Pandas DataFrame containing, trained models with their performance metrics.
-        """ 
-        self._display_msg(msg="Evaluating models performance ...",
-                          progress_bar = self.progress_bar,
-                          show_data=True)
-        # Empty list for storing model performance metrics
-        model_performance_data = []
-
-        # Mapping feature selection methods to corresponding test data
-        feature_selection_to_test_data = {"lasso": test_data[0], 
-                                          "rfe": test_data[1], 
-                                          "pca": test_data[2]}
-
-        # Iterating over models 
-        for index, model_row in model_info.iterrows():
-            # Extracting model name, model id, feature selection method, and model object
-            model_name, model_id, feature_selection, model_object = model_row['Name'], \
-                model_row['Model-ID'], model_row['Feature-Selection'], model_row['obj']
-
-            # Selecting test data based on feature selection method
-            test_set = feature_selection_to_test_data[feature_selection]
-
-            # Model evaluation
-            if model_name == 'knn':
-                performance_metrics = model_object.evaluate(test_data=test_set)
-            else:
-                eval_params = _ModelTraining._eval_params_generation(model_name, 
-                                                                     self.target_column,
-                                                                     self.task_type)
-                performance_metrics = model_object.evaluate(newdata=test_set, **eval_params)
-
-            # Extracting performance metrics
-            if self.is_classification_type():
-                # Classification
-                # Extract performance metrics from the output data
-                performance_metrics_list = [metric[2] for metric in performance_metrics.output_data.itertuples()]
-
-                # Combine all the elements to form a new row
-                new_row = [model_name, model_id, feature_selection] + performance_metrics_list + [model_object]
-            else:
-                # Regression
-                regression_metrics = next(performance_metrics.result.itertuples())
-                sample_size = test_set.select('id').size
-                feature_count = len(test_set.columns) - 2
-                r2_score = regression_metrics[8]
-                adjusted_r2_score = 1 - ((1 - r2_score) * (sample_size - 1) / (sample_size - feature_count - 1))
-                new_row = [model_name, model_id, feature_selection, regression_metrics[0], 
-                           regression_metrics[1], regression_metrics[2], regression_metrics[5], 
-                           regression_metrics[6], r2_score, adjusted_r2_score, model_object]
-
-            model_performance_data.append(new_row)
-
-        if self.is_classification_type():
-            model_metrics_df = pd.DataFrame(model_performance_data, columns=['Name','Model-ID',
-                                                        'Feature-Selection','Accuracy','Micro-Precision',
-                                                        'Micro-Recall','Micro-F1',
-                                                        'Macro-Precision','Macro-Recall',
-                                                        'Macro-F1','Weighted-Precision',
-                                                        'Weighted-Recall','Weighted-F1',
-                                                        'model-obj'])
-        else:
-            model_metrics_df = pd.DataFrame(model_performance_data, columns=['Name', 'Model-ID',
-                                                            'Feature-Selection',
-                                                            'MAE', 'MSE', 'MSLE',
-                                                            'RMSE', 'RMSLE',
-                                                            'R2-score',
-                                                            'Adjusted R2-score',       
-                                                            'model-obj'])
-        self._display_msg(msg="Evaluation completed.",
-                          progress_bar = self.progress_bar,
-                          show_data=True)
-
-        return model_metrics_df
-    
+        return models_df
+  
     def _hyperparameter_tunning(self,
                                 model_param, 
-                                train_data,
-                                test_data):
+                                train_data):
         """
         DESCRIPTION:
             Internal function performs hyperparameter tuning on 
@@ -890,11 +820,6 @@ class _ModelTraining:
                 Specifies the training datasets.
                 Types: tuple of Teradataml DataFrame
             
-            test_data
-                Required Argument.
-                Specifies the testing datasets
-                Types: tuple of Teradataml DataFrame
-            
         RETURNS:
             pandas DataFrame containing, trained models information.
         """ 
@@ -910,13 +835,21 @@ class _ModelTraining:
         # Input columns for model
         model_param['input_columns'] = self.features
 
+        # Setting persist for model
+        model_param['persist'] = self.persist
+
         self._display_msg(msg=model_param['name'], 
                           progress_bar=self.progress_bar,
                           show_data=True)
         
-        # Defining test data for KNN
+        # As we are using entire data for HPT training. So, 
+        # passing prepared training data as test_data for KNN.
         if model_param['name'] == 'knn':
-            model_param['test_data'] = test_data
+            model_param['test_data'] = train_data
+
+        if self.task_type == "Classification":
+            model_param['output_prob'] = True
+            model_param['output_responses'] = self.output_response
 
         # Using RandomSearch for hyperparameter tunning when max_models is given.
         # Otherwise, using GridSearch for hyperparameter tunning.
@@ -951,20 +884,35 @@ class _ModelTraining:
                     sample_id_column='id',stratify_column=self.startify_col, verbose=verbose, max_time=self.max_runtime_secs)
 
         # Getting all passed models
-        _df = _obj.model_stats.merge(_obj.models[_obj.models['STATUS']=='PASS'][['MODEL_ID', 'DATA_ID']], on='MODEL_ID', how='inner')
+        model_info = _obj.model_stats.merge(_obj.models[_obj.models['STATUS']=='PASS'][['MODEL_ID', 'DATA_ID', 'PARAMETERS']],
+                                            on='MODEL_ID', how='inner')
         # Creating mapping data ID to feature selection method
-        data_id_to_method_map = {"DF_0": "lasso", "DF_1": "rfe", "DF_2": "pca"}
-        
-        # Mapping data ID to feature selection method
-        _df['Feature-Selection'] = _df['DATA_ID'].map(data_id_to_method_map)
-        # Getting model details
-        _df['Name'] = model_param['name']
-        _df['Model-ID'] = _df['MODEL_ID']
-        _df['obj'] = _df['MODEL_ID'].apply(lambda x: _obj.get_model(x))
-        
-        # Extracting needed columns
-        model_info = _df[["Name", "Model-ID", "Feature-Selection", "obj"]]
-        
+        data_id_to_table_map = {"DF_0": ('lasso', train_data[0]._table_name), 
+                                "DF_1": ('rfe', train_data[1]._table_name),
+                                "DF_2": ('pca', train_data[2]._table_name)}
+        
+        # Updating model stats with feature selection method and result table
+        for index, row in model_info.iterrows():
+            model_info.loc[index, 'FEATURE_SELECTION'] = data_id_to_table_map[row['DATA_ID']][0]
+            model_info.loc[index, 'DATA_TABLE'] = data_id_to_table_map[row['DATA_ID']][1]
+            model_info.loc[index, 'RESULT_TABLE'] = _obj.get_model(row['MODEL_ID']).result._table_name
+            model_info.loc[index, 'model-obj'] = _obj.get_model(row['MODEL_ID'])
+        
+        # Dropping column 'DATA_ID'
+        model_info.drop(['DATA_ID'], axis=1, inplace=True)
+
+        model_info.insert(1, 'FEATURE_SELECTION', model_info.pop('FEATURE_SELECTION'))
+        
+        if not self.is_classification_type():
+            # Calculating Adjusted-R2 for regression
+            # Getting size and feature count for each feature selection method
+            methods = ["lasso", "rfe", "pca"]
+            size_map = {method : df.select('id').size for method, df in zip(methods, train_data)}
+            feature_count_map = {method : len(df.columns) - 2 for method, df in zip(methods, train_data)}
+            model_info['ADJUSTED_R2'] = model_info.apply(lambda row: 
+                1 - ((1 - row['R2']) * (size_map[row['FEATURE_SELECTION']] - 1) / 
+                (size_map[row['FEATURE_SELECTION']] - feature_count_map[row['FEATURE_SELECTION']] - 1)), axis=1)
+
         self._display_msg(msg="-"*100,
                           progress_bar=self.progress_bar,
                           show_data=True)
@@ -1006,21 +954,36 @@ class _ModelTraining:
         # Setting the eval_params
         eval_params = {"id_column": "id",
                         "accumulate": target_column}
+        
+        model_type = {
+            'xgboost': 'model_type',
+            'glm': 'model_type',
+            'decisionforest': 'tree_type',
+            'svm': 'model_type',
+            'knn': 'model_type'
+        }
+
+        ml_name = ml_name.replace('_', '').lower()
 
         # For Classification
         if task_type.lower() != "regression":
+            eval_params[model_type[ml_name]] = 'Classification'
+            eval_params['output_prob'] = True
+
             if ml_name == 'xgboost':
-                eval_params['model_type'] = 'Classification'
                 eval_params['object_order_column'] = ['task_index', 'tree_num', 'iter','class_num', 'tree_order']
-            else:
-                if ml_name == 'glm':
-                    eval_params['family'] = 'BINOMIAL'
-                    
-                eval_params['output_prob'] = True
+
+            elif ml_name == 'glm':
+                eval_params['family'] = 'BINOMIAL'
+            
         else:
         # For Regression
+            eval_params[model_type[ml_name]] = 'Regression'
+
             if ml_name == 'xgboost':
-                eval_params['model_type'] = 'Regression'
                 eval_params['object_order_column'] = ['task_index', 'tree_num', 'iter', 'tree_order']
+
+            elif ml_name == 'glm':
+                eval_params['family'] = 'GAUSSIAN'
                 
         return eval_params

teradataml/catalog/byom.py

@@ -26,7 +26,7 @@ from teradataml.options.display import display
 from teradataml.common.constants import ModelCatalogingConstants as mac
 from teradataml.options.configure import configure
 from teradataml.utils.utils import execute_sql
-from teradatasqlalchemy.telemetry.queryband import collect_queryband
+from teradataml.telemetry_utils.queryband import collect_queryband
 
 validator = _Validators()
 
@@ -541,13 +541,12 @@ def save_byom(model_id,
         # If exists, extract required information about table columns types
         # else extract from additional_columns_types.
         # Also validate model_id against allowed length.
-        table_exists = connection.dialect.has_table(connection, table_name=table_name, schema=schema_name)
+        table_exists = connection.dialect.has_table(connection, table_name=table_name,
+                                                    schema=schema_name, table_only=True)
         if table_exists:
             # Check if model exists or not. If exists, raise error.
             __check_if_model_exists(
                 model_id, table_name, schema_name, raise_error_if_model_found=True)
-            if len(additional_columns_types) != 0:
-                warnings.warn("Argument additional_columns_types is ignored since table already exists.", stacklevel=2)
 
             # Gather column name and type information from existing table
             existing_table_df = DataFrame(in_schema(schema_name, table_name))
@@ -807,7 +806,7 @@ def delete_byom(model_id, table_name=None, schema_name=None):
 
     # Before proceed further, check whether table exists or not.
     conn = get_connection()
-    if not conn.dialect.has_table(conn, table_name=table_name, schema=schema_name):
+    if not conn.dialect.has_table(conn, table_name=table_name, schema=schema_name, table_only=True):
         error_code = MessageCodes.MODEL_CATALOGING_OPERATION_FAILED
         error_msg = Messages.get_message(
             error_code, "delete", 'Table "{}.{}" does not exist.'.format(schema_name, table_name))
@@ -1472,7 +1471,7 @@ def retrieve_byom(model_id,
 
     # Before proceeding further, check whether table exists or not.
     conn = get_connection()
-    if not conn.dialect.has_table(conn, table_name=table_name, schema=schema_name):
+    if not conn.dialect.has_table(conn, table_name=table_name, schema=schema_name, table_only=True):
         error_code = MessageCodes.MODEL_CATALOGING_OPERATION_FAILED
         error_msg = Messages.get_message(
             error_code, "retrieve", 'Table "{}.{}" does not exist.'.format(schema_name, table_name))
@@ -1535,7 +1534,8 @@ def retrieve_byom(model_id,
     license_table = in_schema(license_schema_name, license_table_name)
 
     # Check whether license table exists or not before proceed further.
-    if not conn.dialect.has_table(conn, table_name=license_table_name, schema=license_schema_name):
+    if not conn.dialect.has_table(conn, table_name=license_table_name, schema=license_schema_name,
+                                  table_only=True):
         error_code = MessageCodes.EXECUTION_FAILED
         error_msg = Messages.get_message(
             error_code, "retrieve the model", 'Table "{}" does not exist.'.format(license_table))
@@ -1723,7 +1723,7 @@ def list_byom(table_name=None, schema_name=None, model_id=None):
 
     # Before proceeding further, check whether table exists or not.
     conn = get_connection()
-    if not conn.dialect.has_table(conn, table_name=table_name, schema=schema_name):
+    if not conn.dialect.has_table(conn, table_name=table_name, schema=schema_name, table_only=True):
         error_code = MessageCodes.MODEL_CATALOGING_OPERATION_FAILED
         error_msg = Messages.get_message(
             error_code, "list", 'Table "{}.{}" does not exist.'.format(schema_name, table_name))

teradataml/clients/pkce_client.py

@@ -425,7 +425,7 @@ class _DAWorkflow:
         """
         device_cfg = requests.post(
             url=self.device_auth_end_point,
-            data={'client_id': self.__client_id})
+            data={'client_id': self.__client_id, 'scope': 'openid'})
 
         # Check the status. If response is not 200, raise error.
         _Validators._validate_http_response(device_cfg, 200, "get the device metadata")

teradataml/common/constants.py

@@ -425,6 +425,8 @@ class TableOperatorConstants(Enum):
     MAP_TEMPLATE = "dataframe_map.template"
     # Template of the intermediate script that will be generated.
     APPLY_TEMPLATE = "dataframe_apply.template"
+    # Template of the intermediate script that will be generated for UDF.
+    UDF_TEMPLATE = "dataframe_udf.template"
     # In-DB execution mode.
     INDB_EXEC = "IN-DB"
     # Local execution mode.
@@ -439,6 +441,8 @@ class TableOperatorConstants(Enum):
     MAP_PARTITION_OP = "map_partition"
     # apply operation.
     APPLY_OP = "apply"
+    # udf operation.
+    UDF_OP = "udf"
     # Template of the script_executor that will be used to generate the temporary script_executor file.
     SCRIPT_TEMPLATE = "script_executor.template"
     # Log Type.
@@ -1369,6 +1373,7 @@ class TeradataAnalyticFunctionTypes(Enum):
     UAF = "UAF"
     TABLEOPERATOR = "TABLE_OPERATOR"
     BYOM = "BYOM"
+    STORED_PROCEDURE = "STORED_PROCEDURE"
 
 
 class TeradataAnalyticFunctionInfo(Enum):
@@ -1379,6 +1384,8 @@ class TeradataAnalyticFunctionInfo(Enum):
     TABLE_OPERATOR = {"func_type": "tableoperator", "lowest_version": "17.00 ",
                       "display_function_type_name" :"TABLE OPERATOR"}
     BYOM = {"func_type": "byom", "lowest_version": None, "display_function_type_name": "BYOM"}
+    STORED_PROCEDURE = {"func_type": "storedprocedure", "lowest_version": "17.20", "display_function_type_name": "UAF",
+           "metadata_class" : "_AnlyFuncMetadataUAF"}
 
 class TeradataUAFSpecificArgs(Enum):
     INPUT_MODE = "input_mode"
@@ -1436,3 +1443,33 @@ class CloudProvider(Enum):
     # and '2018-03-28', using the latest one.
     X_MS_VERSION = "2019-12-12"
     X_MS_BLOB_TYPE = "BlockBlob"
+
+class SessionParamsSQL:
+    # Holds the SQL Statements for Session params.
+    TIMEZONE = "SET TIME ZONE {}"
+    ACCOUNT = "SET SESSION ACCOUNT = '{}' FOR {}"
+    CALENDAR = "SET SESSION CALENDAR = {}"
+    CHARACTER_SET_UNICODE = "SET SESSION CHARACTER SET UNICODE PASS THROUGH {}"
+    COLLATION = "SET SESSION COLLATION {}"
+    CONSTRAINT = "SET SESSION CONSTRAINT = {}"
+    DATABASE = "SET SESSION DATABASE {}"
+    DATEFORM = "SET SESSION DATEFORM = {}"
+    DEBUG_FUNCTION = "SET SESSION DEBUG FUNCTION {} {}"
+    DOT_NOTATION = "SET SESSION DOT NOTATION {} ON ERROR"
+    ISOLATED_LOADING = "SET SESSION FOR {} ISOLATED LOADING"
+    FUNCTION_TRACE = "SET SESSION FUNCTION TRACE USING {} FOR TABLE {}"
+    JSON_IGNORE_ERRORS = "SET SESSION JSON IGNORE ERRORS {}"
+    SEARCHUIFDBPATH = "SET SESSION SEARCHUIFDBPATH = {}"
+    TRANSACTION_ISOLATION_LEVEL = "SET SESSION CHARACTERISTICS AS TRANSACTION ISOLATION LEVEL {}"
+    QUERY_BAND = "SET QUERY_BAND = {} FOR {}"
+    UDFSEARCHPATH = "SET SESSION UDFSEARCHPATH = {} FOR FUNCTION = {}"
+
+class SessionParamsPythonNames:
+    # Holds the SQL Statements for Session params.
+    TIMEZONE = "Session Time Zone"
+    ACCOUNT = "Account Name"
+    CALENDAR = "Calendar"
+    COLLATION = "Collation"
+    DATABASE = "Current DataBase"
+    DATEFORM = 'Current DateForm'
+