Merge remote-tracking branch 'origin/development' into sweep

pyproject.toml

@@ -17,7 +17,7 @@ ingester3 = "^2.1.0"
 properscoring = "^0.1"
 wandb = "^0.18.7"
 pyprojroot="^0.3.0"
-views_evaluation = { git = "https://github.com/views-platform/views-evaluation.git", branch = "main" }
+views-evaluation = { git = "https://github.com/views-platform/views-evaluation.git", branch = "main" }
 views-forecasts = "^0.5.5"
 
 

tests/test_model_manager.py

@@ -89,7 +89,8 @@ def get_deployment_config():
                 manager._wandb_alert(title="Test Alert", text="This is a test alert", level="info")
                 mock_alert.assert_called_once_with(title="Test Alert", text="This is a test alert", level="info")
 
-def test_model_manager_init(mock_model_path):
+@patch("views_forecasts.extensions.ForecastAccessor.read_store")
+def test_model_manager_init(mock_model_path, mock_read_store):
     """
     Test the initialization of the ModelManager class.
     
@@ -133,7 +134,8 @@ def get_meta_config():
         assert manager._config_hyperparameters == {"hp_key": "hp_value"}
         assert manager._config_meta == {"meta_key": "meta_value"}
 
-def test_load_config(mock_model_path):
+@patch("views_forecasts.extensions.ForecastAccessor.read_store")
+def test_load_config(mock_model_pat, mock_read_store):
     """
     Test the __load_config method of the ModelManager class.
     

views_pipeline_core/managers/ensemble.py

@@ -5,7 +5,7 @@
 from views_pipeline_core.files.utils import save_dataframe, read_dataframe
 from views_pipeline_core.configs.pipeline import PipelineConfig
 from views_evaluation.evaluation.metrics import MetricsManager
-from views_forecasts.extensions import *
+from views_forecasts.extensions import * 
 from typing import Union, Optional, List, Dict
 import wandb
 import logging
@@ -353,6 +353,7 @@ def _train_model_artifact(
     def _evaluate_model_artifact(
         self, model_name: str, run_type: str, eval_type: str
     ) -> List[pd.DataFrame]:
+        # from views_forecasts.extensions import ForecastAccessor
         logger.info(f"Evaluating single model {model_name}...")
 
         model_path = ModelPathManager(model_name)
@@ -423,6 +424,7 @@ def _evaluate_model_artifact(
         return preds
 
     def _forecast_model_artifact(self, model_name: str, run_type: str) -> pd.DataFrame:
+        # from views_forecasts.extensions import ForecastAccessor
         logger.info(f"Forecasting single model {model_name}...")
 
         model_path = ModelPathManager(model_name)

views_pipeline_core/managers/model.py

@@ -25,7 +25,6 @@
 from views_evaluation.evaluation.metrics import MetricsManager
 from views_forecasts.extensions import *
 import traceback
-import numpy as np
 
 
 logger = logging.getLogger(__name__)
@@ -648,9 +647,9 @@ def __init__(
     ) -> None:
         """
         Initializes the ModelManager with the given model path.
-
         Args:
             model_path (ModelPathManager): The path manager for the model.
+            wandb_notifications (bool, optional): Whether to enable WandB notifications. Defaults to True.
         """
         self._entity = "views_pipeline"
         self._model_repo = "views-models"
@@ -786,7 +785,8 @@ def __get_pred_store_name(self) -> str:
         #     self._owner, self._model_repo
         # )
         from views_pipeline_core.managers.package import PackageManager
-
+        # from views_forecasts.extensions import ViewsMetadata
+
         version = PackageManager.get_latest_release_version_from_github(
             repository_name=self._model_repo
         )
@@ -937,18 +937,20 @@ def _generate_evaluation_table(self, metric_dict: Dict) -> str:
         Returns:
             str: A formatted string representing the evaluation table.
         """
-        table_str = "\n{:<70} {:<30}".format("Metric", "Value")
-        table_str += "-" * 100 + "\n"
+        from tabulate import tabulate
+        # create an empty dataframe with columns 'Metric' and 'Value'
+        metric_df = pd.DataFrame(columns=["Metric", "Value"])
         for key, value in metric_dict.items():
             try:
                 if not str(key).startswith("_"):
-                    value = float(
-                        value
-                    )  # Super hacky way to filter out metrics. 0/10 do not recommend
-                    table_str += "{:<70}\t{:<30}\n".format(str(key), value)
+                    value = float(value)
+                    # add metric and value to the dataframe
+                    metric_df = metric_df.append({"Metric": key, "Value": value}, ignore_index=True)
             except:
                 continue
-        return table_str
+        result = tabulate(metric_df, headers='keys', tablefmt='grid')
+        print(result)
+        return f"```\n{result}\n```"
 
     def _save_model_artifact(self, run_type: str) -> None:
         """
@@ -1085,6 +1087,7 @@ def _save_predictions(
             path_generated (str or Path): The path where the predictions should be saved.
             sequence_number (int): The sequence number.
         """
+        from views_forecasts.extensions import ForecastAccessor
         try:
             path_generated = Path(path_generated)
             path_generated.mkdir(parents=True, exist_ok=True)