featrixsphere 0.2.5182__tar.gz → 0.2.5183__tar.gz

{featrixsphere-0.2.5182 → featrixsphere-0.2.5183}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: featrixsphere
-Version: 0.2.5182
+Version: 0.2.5183
 Summary: Transform any CSV into a production-ready ML model in minutes, not months.
 Home-page: https://github.com/Featrix/sphere
 Author: Featrix

{featrixsphere-0.2.5182 → featrixsphere-0.2.5183}/featrixsphere/__init__.py

@@ -38,7 +38,7 @@ Example:
     ...                                labels=['Experiment A', 'Experiment B'])
 """
 
-__version__ = "0.2.5182"
+__version__ = "0.2.5183"
 __author__ = "Featrix"
 __email__ = "support@featrix.com"
 __license__ = "MIT"

{featrixsphere-0.2.5182 → featrixsphere-0.2.5183}/featrixsphere/client.py

@@ -8287,6 +8287,55 @@ class FeatrixSphereClient:
         """
         return PredictionGrid(session_id, self, degrees_of_freedom, grid_shape, target_column)
 
+    def get_embedding_space_columns(self, session_id: str) -> Dict[str, Any]:
+        """
+        Get column names and types from the embedding space.
+        
+        Tries to get from model_card.json first (if training completed), 
+        otherwise falls back to loading the embedding space directly.
+        
+        Args:
+            session_id: Session ID with trained embedding space
+            
+        Returns:
+            Dictionary with:
+                - column_names: List of column names
+                - column_types: Dict mapping column names to types (scalar, set, free_string, etc.)
+                - num_columns: Total number of columns
+                
+        Example:
+            >>> columns = client.get_embedding_space_columns(session_id)
+            >>> print(f"Columns: {columns['column_names']}")
+            >>> print(f"Types: {columns['column_types']}")
+        """
+        # Try model_card first (if training completed)
+        try:
+            model_card = self.get_model_card(session_id)
+            
+            # Extract column names from training_dataset.feature_names
+            training_dataset = model_card.get('training_dataset', {})
+            column_names = training_dataset.get('feature_names', [])
+            
+            # Extract column types from feature_inventory
+            feature_inventory = model_card.get('feature_inventory', {})
+            column_types = {}
+            for feature_name, feature_info in feature_inventory.items():
+                if isinstance(feature_info, dict):
+                    column_types[feature_name] = feature_info.get('type', 'unknown')
+            
+            if column_names:
+                return {
+                    "column_names": column_names,
+                    "column_types": column_types,
+                    "num_columns": len(column_names)
+                }
+        except Exception:
+            # Model card doesn't exist yet, fall back to direct endpoint
+            pass
+        
+        # Fallback: load embedding space directly
+        return self._get_json(f"/compute/session/{session_id}/columns")
+
 
 class PredictionGrid:
     """
@@ -9032,55 +9081,6 @@ class PredictionGrid:
         else:
             return fig
     
-    def get_embedding_space_columns(self, session_id: str) -> Dict[str, Any]:
-        """
-        Get column names and types from the embedding space.
-        
-        Tries to get from model_card.json first (if training completed), 
-        otherwise falls back to loading the embedding space directly.
-        
-        Args:
-            session_id: Session ID with trained embedding space
-            
-        Returns:
-            Dictionary with:
-                - column_names: List of column names
-                - column_types: Dict mapping column names to types (scalar, set, free_string, etc.)
-                - num_columns: Total number of columns
-                
-        Example:
-            >>> columns = client.get_embedding_space_columns(session_id)
-            >>> print(f"Columns: {columns['column_names']}")
-            >>> print(f"Types: {columns['column_types']}")
-        """
-        # Try model_card first (if training completed)
-        try:
-            model_card = self.get_model_card(session_id)
-            
-            # Extract column names from training_dataset.feature_names
-            training_dataset = model_card.get('training_dataset', {})
-            column_names = training_dataset.get('feature_names', [])
-            
-            # Extract column types from feature_inventory
-            feature_inventory = model_card.get('feature_inventory', {})
-            column_types = {}
-            for feature_name, feature_info in feature_inventory.items():
-                if isinstance(feature_info, dict):
-                    column_types[feature_name] = feature_info.get('type', 'unknown')
-            
-            if column_names:
-                return {
-                    "column_names": column_names,
-                    "column_types": column_types,
-                    "num_columns": len(column_names)
-                }
-        except Exception:
-            # Model card doesn't exist yet, fall back to direct endpoint
-            pass
-        
-        # Fallback: load embedding space directly
-        return self._get_json(f"/compute/session/{session_id}/columns")
-    
     def get_predictor_schema(self, session_id: str, predictor_index: int = 0) -> Dict[str, Any]:
         """
         Get predictor schema/metadata for validating input data locally.

{featrixsphere-0.2.5182 → featrixsphere-0.2.5183}/featrixsphere.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: featrixsphere
-Version: 0.2.5182
+Version: 0.2.5183
 Summary: Transform any CSV into a production-ready ML model in minutes, not months.
 Home-page: https://github.com/Featrix/sphere
 Author: Featrix