featrixsphere 0.2.1237__tar.gz → 0.2.1314__tar.gz

{featrixsphere-0.2.1237/featrixsphere.egg-info → featrixsphere-0.2.1314}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: featrixsphere
-Version: 0.2.1237
+Version: 0.2.1314
 Summary: Transform any CSV into a production-ready ML model in minutes, not months.
 Home-page: https://github.com/Featrix/sphere
 Author: Featrix

featrixsphere-0.2.1314/VERSION

@@ -0,0 +1 @@
+0.2.1314

{featrixsphere-0.2.1237 → featrixsphere-0.2.1314}/featrixsphere/__init__.py

@@ -38,7 +38,7 @@ Example:
     ...                                labels=['Experiment A', 'Experiment B'])
 """
 
-__version__ = "0.2.1237"
+__version__ = "0.2.1314"
 __author__ = "Featrix"
 __email__ = "support@featrix.com"
 __license__ = "MIT"

{featrixsphere-0.2.1237 → featrixsphere-0.2.1314}/featrixsphere/client.py

@@ -1727,48 +1727,6 @@ class FeatrixSphereClient:
     # File Upload
     # =========================================================================
     
-    def upload_file(self, file_path: str) -> str:
-        """
-        Upload a file to the server without creating a session.
-        Returns the filename that can be used in training requests.
-        
-        Args:
-            file_path: Path to the file to upload
-            
-        Returns:
-            Filename (relative path) that can be used in training requests
-        """
-        from pathlib import Path as PathLib
-        file_path_obj = PathLib(file_path)
-        if not file_path_obj.exists():
-            raise FileNotFoundError(f"File not found: {file_path}")
-        
-        with open(file_path_obj, 'rb') as f:
-            files = {'file': (file_path_obj.name, f, 'text/csv' if file_path_obj.suffix == '.csv' else 'application/gzip')}
-            response = self._make_request("POST", "/compute/upload_file", files=files)
-        
-        response_data = response.json()
-        
-        # Handle S3 upload response (returns s3_url and filename)
-        if 's3_url' in response_data:
-            # S3 upload - extract filename from key or use returned filename
-            filename = response_data.get('filename')
-            if not filename:
-                # Extract from S3 key if filename not provided
-                s3_key = response_data.get('key', '')
-                if s3_key:
-                    filename = PathLib(s3_key).name
-            if not filename:
-                raise ValueError("Server did not return filename in S3 upload response")
-            return filename
-        
-        # Handle local file upload response (returns filename)
-        filename = response_data.get('filename')
-        if not filename:
-            raise ValueError("Server did not return filename in upload response")
-        
-        return filename
-
     def upload_file_and_create_session(self, file_path: Path, session_name_prefix: str = None, name: str = None, webhooks: Dict[str, str] = None) -> SessionInfo:
         """
         Upload a CSV, Parquet, JSON, or JSONL file and create a new session.
@@ -4159,7 +4117,6 @@ class FeatrixSphereClient:
                                     epochs: int = 0, 
                                     rare_label_value: str = None,
                                     class_imbalance: dict = None,
-                                    optimize_for: str = "balanced",
                                     poll_interval: int = 30, max_poll_time: int = 3600,
                                     verbose: bool = True,
                                     webhooks: Dict[str, str] = None) -> SessionInfo:
@@ -4182,7 +4139,6 @@ class FeatrixSphereClient:
             epochs: Number of training epochs (default: 0; automatic)
             rare_label_value: For binary classification, which class is the rare/minority class for metrics (default: None)
             class_imbalance: Expected class ratios/counts from real world for sampled data (default: None)
-            optimize_for: Optimization target - "balanced" (F1 score), "precision", or "recall" (default: "balanced")
             poll_interval: Seconds between status checks when job is already running (default: 30)
             max_poll_time: Maximum time to poll in seconds (default: 3600 = 1 hour)
             verbose: Whether to print status updates during polling (default: True)
@@ -4220,7 +4176,6 @@ class FeatrixSphereClient:
                 "target_column": target_column,
                 "target_column_type": target_column_type,
                 "epochs": str(epochs),
-                "optimize_for": optimize_for,
             }
             
             # Handle file upload - send file directly in multipart form
@@ -4339,7 +4294,6 @@ class FeatrixSphereClient:
                               validation_ignore_columns: List[str] = None,
                               rare_label_value: str = None,
                               class_imbalance: dict = None,
-                              optimize_for: str = "balanced",
                               cost_false_positive: float = None,
                               cost_false_negative: float = None,
                               poll_interval: int = 30, max_poll_time: int = 3600,
@@ -4519,24 +4473,6 @@ class FeatrixSphereClient:
         This happens automatically - no configuration needed. The system invests a few seconds
         in analysis to deliver significantly better models.
         
-        Understanding optimize_for:
-        ---------------------------
-        The optimize_for parameter controls which loss function and training strategy is used,
-        optimizing for different aspects of model performance:
-        
-        - "balanced" (default): Optimizes for F1 score (harmonic mean of precision and recall).
-          Uses FocalLoss with class weights. Best for general-purpose classification where you
-          want balanced performance across all classes.
-          
-        - "precision": Optimizes for precision (minimizing false positives). Uses FocalLoss with
-          class weights, which focuses training on hard-to-classify examples. Best when false
-          positives are costly (e.g., fraud detection where flagging legitimate transactions
-          as fraud is expensive).
-          
-        - "recall": Optimizes for recall (minimizing false negatives). Uses CrossEntropyLoss
-          with class weights that strongly boost the minority class. Best when false negatives
-          are costly (e.g., medical diagnosis where missing a disease is dangerous).
-        
         Understanding class_imbalance:
         ------------------------------
         For imbalanced datasets where your training data doesn't reflect real-world class
@@ -4657,7 +4593,6 @@ class FeatrixSphereClient:
             target_column='approved',
             target_column_type='set',
             class_imbalance={'approved': 0.97, 'rejected': 0.03},
-            optimize_for='recall',  # Don't miss rejections
             rare_label_value='rejected'
         )
         
@@ -4678,7 +4613,6 @@ class FeatrixSphereClient:
             target_column='is_fraud',
             target_column_type='set',
             rare_label_value='fraud',
-            optimize_for='precision',  # Minimize false alarms
             class_imbalance={'legitimate': 0.999, 'fraud': 0.001}
         )
         ```
@@ -4692,8 +4626,7 @@ class FeatrixSphereClient:
             session_id=session.session_id,
             target_column='has_disease',
             target_column_type='set',
-            rare_label_value='positive',
-            optimize_for='recall'  # Don't miss any cases
+            rare_label_value='positive'
         )
         ```
         
@@ -4735,14 +4668,6 @@ class FeatrixSphereClient:
             validation_ignore_columns: List of column names to exclude from validation queries (default: None)
             rare_label_value: For binary classification, which class is the rare/minority class for metrics (default: None)
             class_imbalance: Expected class ratios/counts from real world for sampled data (default: None)
-            optimize_for: Optimization target - "balanced" (F1 score), "precision", or "recall" (default: "balanced").
-                Ignored if cost_false_positive and cost_false_negative are provided.
-            cost_false_positive: Cost of a false positive (predicting positive when actually negative). 
-                Must be specified together with cost_false_negative. Only valid for target_column_type="set".
-                When provided, overrides optimize_for and uses cost-based optimization.
-            cost_false_negative: Cost of a false negative (predicting negative when actually positive).
-                Must be specified together with cost_false_positive. Only valid for target_column_type="set".
-                When provided, overrides optimize_for and uses cost-based optimization.
             poll_interval: Seconds between status checks when job is already running (default: 30)
             max_poll_time: Maximum time to poll in seconds (default: 3600 = 1 hour)
             verbose: Whether to print status updates during polling (default: True)
@@ -4769,7 +4694,6 @@ class FeatrixSphereClient:
                 raise ValueError("cost_false_positive and cost_false_negative must be positive numbers")
             if verbose:
                 print(f"💰 Cost-based optimization enabled: FP cost={cost_false_positive}, FN cost={cost_false_negative}")
-                print(f"   (optimize_for='{optimize_for}' will be ignored)")
         
         # If DataFrame provided, save to temp file and use file_path logic
         temp_file = None
@@ -4803,7 +4727,6 @@ class FeatrixSphereClient:
                     epochs=epochs,
                     rare_label_value=rare_label_value,
                     class_imbalance=class_imbalance,
-                    optimize_for=optimize_for,
                     cost_false_positive=cost_false_positive,
                     cost_false_negative=cost_false_negative,
                     verbose=verbose,
@@ -4817,8 +4740,7 @@ class FeatrixSphereClient:
                 "epochs": epochs,
                 "validation_ignore_columns": validation_ignore_columns or [],
                 "rare_label_value": rare_label_value,
-                "class_imbalance": class_imbalance,
-                "optimize_for": optimize_for
+                "class_imbalance": class_imbalance
             }
             if cost_false_positive is not None and cost_false_negative is not None:
                 data["cost_false_positive"] = cost_false_positive
@@ -5395,7 +5317,6 @@ class FeatrixSphereClient:
         epochs: int,
         rare_label_value: str,
         class_imbalance: dict,
-        optimize_for: str,
         cost_false_positive: float = None,
         cost_false_negative: float = None,
         verbose: bool = True,
@@ -5423,8 +5344,7 @@ class FeatrixSphereClient:
         data = {
             'target_column': target_column,
             'target_column_type': target_column_type,
-            'epochs': str(epochs),
-            'optimize_for': optimize_for,
+            'epochs': str(epochs)
         }
         
         if rare_label_value:

{featrixsphere-0.2.1237 → featrixsphere-0.2.1314/featrixsphere.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: featrixsphere
-Version: 0.2.1237
+Version: 0.2.1314
 Summary: Transform any CSV into a production-ready ML model in minutes, not months.
 Home-page: https://github.com/Featrix/sphere
 Author: Featrix

featrixsphere-0.2.1237/VERSION

@@ -1 +0,0 @@
-0.2.1237