dragon-ml-toolbox 13.3.0__py3-none-any.whl → 16.2.0__py3-none-any.whl

ml_tools/ML_vision_transformers.py

@@ -0,0 +1,284 @@
+from typing import Union, Dict, Type, Callable, Optional, Any, List, Literal
+from PIL import ImageOps, Image
+from torchvision import transforms
+from pathlib import Path
+import json
+
+from ._logger import _LOGGER
+from ._script_info import _script_info
+from ._keys import VisionTransformRecipeKeys
+from .path_manager import make_fullpath
+
+
+__all__ = [
+    "TRANSFORM_REGISTRY",
+    "ResizeAspectFill",
+    "create_offline_augmentations"
+]
+
+# --- Custom Vision Transform Class ---
+class ResizeAspectFill:
+    """
+    Custom transformation to make an image square by padding it to match the
+    longest side, preserving the aspect ratio. The image is finally centered.
+
+    Args:
+        pad_color (Union[str, int]): Color to use for the padding.
+                                     Defaults to "black".
+    """
+    def __init__(self, pad_color: Union[str, int] = "black") -> None:
+        self.pad_color = pad_color
+        # Store kwargs to allow for re-creation
+        self.__setattr__(VisionTransformRecipeKeys.KWARGS, {"pad_color": pad_color})
+
+    def __call__(self, image: Image.Image) -> Image.Image:
+        if not isinstance(image, Image.Image):
+            _LOGGER.error(f"Expected PIL.Image.Image, got {type(image).__name__}")
+            raise TypeError()
+
+        w, h = image.size
+        if w == h:
+            return image
+
+        # Determine padding to center the image
+        if w > h:
+            top_padding = (w - h) // 2
+            bottom_padding = w - h - top_padding
+            padding = (0, top_padding, 0, bottom_padding)
+        else: # h > w
+            left_padding = (h - w) // 2
+            right_padding = h - w - left_padding
+            padding = (left_padding, 0, right_padding, 0)
+
+        return ImageOps.expand(image, padding, fill=self.pad_color)
+
+
+#############################################################
+#NOTE: Add custom transforms.
+TRANSFORM_REGISTRY: Dict[str, Type[Callable]] = {
+    "ResizeAspectFill": ResizeAspectFill,
+}
+#############################################################
+
+
+def create_offline_augmentations(
+    input_directory: Union[str, Path],
+    output_directory: Union[str, Path],
+    results_per_image: int,
+    recipe: Optional[Dict[str, Any]] = None,
+    save_format: Literal["WEBP", "JPEG", "PNG", "BMP", "TIF"] = "WEBP",
+    save_quality: int = 80
+) -> None:
+    """
+    Reads all valid images from an input directory, applies augmentations,
+    and saves the new images to an output directory (offline augmentation).
+
+    Skips subdirectories in the input path.
+
+    Args:
+        input_directory (Union[str, Path]): Path to the directory of source images.
+        output_directory (Union[str, Path]): Path to save the augmented images.
+        results_per_image (int): The number of augmented versions to create
+                                 for each source image.
+        recipe (Optional[Dict[str, Any]]): A transform recipe dictionary. If None,
+                                           a default set of strong, random
+                                           augmentations will be used.
+        save_format (str): The format to save images (e.g., "WEBP", "JPEG", "PNG").
+                           Defaults to "WEBP" for good compression.
+        save_quality (int): The quality for lossy formats (1-100). Defaults to 80.
+    """
+    VALID_IMG_EXTENSIONS = ('.jpg', '.jpeg', '.png', '.bmp', '.webp', '.tif', '.tiff')
+    
+    # --- 1. Validate Paths ---
+    in_path = make_fullpath(input_directory, enforce="directory")
+    out_path = make_fullpath(output_directory, make=True, enforce="directory")
+    
+    _LOGGER.info(f"Starting offline augmentation:\n\tInput: {in_path}\n\tOutput: {out_path}")
+
+    # --- 2. Find Images ---
+    image_files = [
+        f for f in in_path.iterdir()
+        if f.is_file() and f.suffix.lower() in VALID_IMG_EXTENSIONS
+    ]
+    
+    if not image_files:
+        _LOGGER.warning(f"No valid image files found in {in_path}.")
+        return
+
+    _LOGGER.info(f"Found {len(image_files)} images to process.")
+
+    # --- 3. Define Transform Pipeline ---
+    transform_pipeline: transforms.Compose
+    
+    if recipe:
+        _LOGGER.info("Building transformations from provided recipe.")
+        try:
+            transform_pipeline = _build_transform_from_recipe(recipe)
+        except Exception as e:
+            _LOGGER.error(f"Failed to build transform from recipe: {e}")
+            return
+    else:
+        _LOGGER.info("No recipe provided. Using default random augmentation pipeline.")
+        # Default "random" pipeline
+        transform_pipeline = transforms.Compose([
+            transforms.RandomResizedCrop(256, scale=(0.4, 1.0)),
+            transforms.RandomHorizontalFlip(p=0.5),
+            transforms.RandomRotation(degrees=90),
+            transforms.ColorJitter(brightness=0.3, contrast=0.3, saturation=0.3, hue=0.15),
+            transforms.RandomPerspective(distortion_scale=0.2, p=0.4),
+            transforms.RandomAffine(degrees=0, translate=(0.1, 0.1)),
+            transforms.RandomApply([
+                transforms.GaussianBlur(kernel_size=3)
+            ], p=0.3)
+        ])
+
+    # --- 4. Process Images ---
+    total_saved = 0
+    format_upper = save_format.upper()
+    
+    for img_path in image_files:
+        _LOGGER.debug(f"Processing {img_path.name}...")
+        try:
+            original_image = Image.open(img_path).convert("RGB")
+            
+            for i in range(results_per_image):
+                new_stem = f"{img_path.stem}_aug_{i+1:03d}"
+                output_path = out_path / f"{new_stem}.{format_upper.lower()}"
+                
+                # Apply transform
+                transformed_image = transform_pipeline(original_image)
+                
+                # Save
+                transformed_image.save(
+                    output_path, 
+                    format=format_upper, 
+                    quality=save_quality,
+                    optimize=True # Add optimize flag
+                )
+                total_saved += 1
+                
+        except Exception as e:
+            _LOGGER.warning(f"Failed to process or save augmentations for {img_path.name}: {e}")
+
+    _LOGGER.info(f"Offline augmentation complete. Saved {total_saved} new images.")
+
+
+def _build_transform_from_recipe(recipe: Dict[str, Any]) -> transforms.Compose:
+    """Internal helper to build a transform pipeline from a recipe dict."""
+    pipeline_steps: List[Callable] = []
+    
+    if VisionTransformRecipeKeys.PIPELINE not in recipe:
+        _LOGGER.error("Recipe dict is invalid: missing 'pipeline' key.")
+        raise ValueError("Invalid recipe format.")
+
+    for step in recipe[VisionTransformRecipeKeys.PIPELINE]:
+        t_name = step.get(VisionTransformRecipeKeys.NAME)
+        t_kwargs = step.get(VisionTransformRecipeKeys.KWARGS, {})
+        
+        if not t_name:
+            _LOGGER.error(f"Invalid transform step, missing 'name': {step}")
+            continue
+        
+        transform_class: Any = None
+
+        # 1. Check standard torchvision transforms
+        if hasattr(transforms, t_name):
+            transform_class = getattr(transforms, t_name)
+        # 2. Check custom transforms
+        elif t_name in TRANSFORM_REGISTRY:
+            transform_class = TRANSFORM_REGISTRY[t_name]
+        # 3. Not found
+        else:
+            _LOGGER.error(f"Unknown transform '{t_name}' in recipe. Not found in torchvision.transforms or TRANSFORM_REGISTRY.")
+            raise ValueError(f"Unknown transform name: {t_name}")
+            
+        # Instantiate the transform
+        try:
+            pipeline_steps.append(transform_class(**t_kwargs))
+        except Exception as e:
+            _LOGGER.error(f"Failed to instantiate transform '{t_name}' with kwargs {t_kwargs}: {e}")
+            raise
+            
+    return transforms.Compose(pipeline_steps)
+
+
+def _save_recipe(recipe: Dict[str, Any], filepath: Path) -> None:
+    """
+    Saves a transform recipe dictionary to a JSON file.
+
+    Args:
+        recipe (Dict[str, Any]): The recipe dictionary to save.
+        filepath (str): The path to the output .json file.
+    """
+    final_filepath = filepath.with_suffix(".json")
+    
+    try:
+        with open(final_filepath, 'w') as f:
+            json.dump(recipe, f, indent=4)
+        _LOGGER.info(f"Transform recipe saved as '{final_filepath.name}'.")
+    except Exception as e:
+        _LOGGER.error(f"Failed to save recipe to '{final_filepath}': {e}")
+        raise
+
+
+def _load_recipe_and_build_transform(filepath: Union[str,Path]) -> transforms.Compose:
+    """
+    Loads a transform recipe from a .json file and reconstructs the
+    torchvision.transforms.Compose pipeline.
+
+    Args:
+        filepath (str): Path to the saved transform recipe .json file.
+
+    Returns:
+        transforms.Compose: The reconstructed transformation pipeline.
+        
+    Raises:
+        ValueError: If a transform name in the recipe is not found in
+                    torchvision.transforms or the custom TRANSFORM_REGISTRY.
+    """
+    # validate filepath
+    final_filepath = make_fullpath(filepath, enforce="file")
+    
+    try:
+        with open(final_filepath, 'r') as f:
+            recipe = json.load(f)
+    except Exception as e:
+        _LOGGER.error(f"Failed to load recipe from '{final_filepath}': {e}")
+        raise
+        
+    pipeline_steps: List[Callable] = []
+    
+    if VisionTransformRecipeKeys.PIPELINE not in recipe:
+        _LOGGER.error("Recipe file is invalid: missing 'pipeline' key.")
+        raise ValueError("Invalid recipe format.")
+
+    for step in recipe[VisionTransformRecipeKeys.PIPELINE]:
+        t_name = step[VisionTransformRecipeKeys.NAME]
+        t_kwargs = step[VisionTransformRecipeKeys.KWARGS]
+        
+        transform_class: Any = None
+
+        # 1. Check standard torchvision transforms
+        if hasattr(transforms, t_name):
+            transform_class = getattr(transforms, t_name)
+        # 2. Check custom transforms
+        elif t_name in TRANSFORM_REGISTRY:
+            transform_class = TRANSFORM_REGISTRY[t_name]
+        # 3. Not found
+        else:
+            _LOGGER.error(f"Unknown transform '{t_name}' in recipe. Not found in torchvision.transforms or TRANSFORM_REGISTRY.")
+            raise ValueError(f"Unknown transform name: {t_name}")
+            
+        # Instantiate the transform
+        try:
+            pipeline_steps.append(transform_class(**t_kwargs))
+        except Exception as e:
+            _LOGGER.error(f"Failed to instantiate transform '{t_name}' with kwargs {t_kwargs}: {e}")
+            raise
+            
+    _LOGGER.info(f"Successfully loaded and built transform pipeline from '{final_filepath.name}'.")
+    return transforms.Compose(pipeline_steps)
+
+
+def info():
+    _script_info(__all__)

ml_tools/PSO_optimization.py

@@ -12,9 +12,9 @@ from .serde import deserialize_object
 from .math_utilities import threshold_binary_values, threshold_binary_values_batch
 from .path_manager import sanitize_filename, make_fullpath, list_files_by_extension
 from ._logger import _LOGGER
-from .keys import EnsembleKeys
+from ._keys import EnsembleKeys
 from ._script_info import _script_info
-from .SQL import DatabaseManager
+from .SQL import DragonSQL
 from .optimization_tools import _save_result
 
 """
@@ -191,7 +191,7 @@ def _set_feature_names(size: int, names: Union[list[str], None]):
         return names
 
 
-def _run_single_pso(objective_function: ObjectiveFunction, pso_args: dict, feature_names: list[str], target_name: str, random_state: int, save_format: Literal['csv', 'sqlite', 'both'], csv_path: Path, db_manager: Optional[DatabaseManager], db_table_name: str):
+def _run_single_pso(objective_function: ObjectiveFunction, pso_args: dict, feature_names: list[str], target_name: str, random_state: int, save_format: Literal['csv', 'sqlite', 'both'], csv_path: Path, db_manager: Optional[DragonSQL], db_table_name: str):
     """Helper for a single PSO run that also handles saving."""
     pso_args.update({"seed": random_state})
     
@@ -213,7 +213,7 @@ def _run_single_pso(objective_function: ObjectiveFunction, pso_args: dict, featu
     return best_features_named, best_target_named
 
 
-def _run_post_hoc_pso(objective_function: ObjectiveFunction, pso_args: dict, feature_names: list[str], target_name: str, repetitions: int, save_format: Literal['csv', 'sqlite', 'both'], csv_path: Path, db_manager: Optional[DatabaseManager], db_table_name: str):
+def _run_post_hoc_pso(objective_function: ObjectiveFunction, pso_args: dict, feature_names: list[str], target_name: str, repetitions: int, save_format: Literal['csv', 'sqlite', 'both'], csv_path: Path, db_manager: Optional[DragonSQL], db_table_name: str):
     """Helper for post-hoc analysis that saves results incrementally."""
     progress = trange(repetitions, desc="Post-Hoc PSO", unit="run")
     for _ in progress:
@@ -342,7 +342,7 @@ def run_pso(lower_boundaries: list[float],
         schema = {"result_id": "INTEGER PRIMARY KEY AUTOINCREMENT", **schema}
         
         # Create table
-        with DatabaseManager(db_path) as db:
+        with DragonSQL(db_path) as db:
             db.create_table(db_table_name, schema)
         
     pso_arguments = {
@@ -357,7 +357,7 @@ def run_pso(lower_boundaries: list[float],
     
     # --- Dispatcher ---
     # Use a real or dummy context manager to handle the DB connection cleanly
-    db_context = DatabaseManager(db_path) if save_format in ['sqlite', 'both'] else nullcontext()
+    db_context = DragonSQL(db_path) if save_format in ['sqlite', 'both'] else nullcontext()
 
     with db_context as db_manager:
         if post_hoc_analysis is None or post_hoc_analysis <= 1:

ml_tools/SQL.py

@@ -9,11 +9,11 @@ from .path_manager import make_fullpath, sanitize_filename
 
 
 __all__ = [
-    "DatabaseManager",
+    "DragonSQL",
 ]
 
 
-class DatabaseManager:
+class DragonSQL:
     """
     A user-friendly context manager for handling SQLite database operations.
 
@@ -35,7 +35,7 @@ class DatabaseManager:
     ...     "feature_a": "REAL",
     ...     "score": "REAL"
     ... }
-    >>> with DatabaseManager("my_results.db") as db:
+    >>> with DragonSQL("my_results.db") as db:
     ...     db.create_table("experiments", schema)
     ...     data = {"run_name": "first_run", "feature_a": 0.123, "score": 95.5}
     ...     db.insert_row("experiments", data)
@@ -43,7 +43,7 @@ class DatabaseManager:
     ...     print(df)
     """
     def __init__(self, db_path: Union[str, Path]):
-        """Initializes the DatabaseManager with the path to the database file."""
+        """Initializes the DragonSQL with the path to the database file."""
         if isinstance(db_path, str):
             if not db_path.endswith(".db"):
                 db_path = db_path + ".db"

ml_tools/_keys.py

@@ -0,0 +1,171 @@
+class MagicWords:
+    """General purpose keys"""
+    LATEST = "latest"
+    CURRENT = "current"
+    RENAME = "rename"
+
+
+class PyTorchLogKeys:
+    """
+    Used internally for ML scripts module.
+    
+    Centralized keys for logging and history.
+    """
+    # --- Epoch Level ---
+    TRAIN_LOSS = 'train_loss'
+    VAL_LOSS = 'val_loss'
+    LEARNING_RATE = 'lr'
+
+    # --- Batch Level ---
+    BATCH_LOSS = 'loss'
+    BATCH_INDEX = 'batch'
+    BATCH_SIZE = 'size'
+
+
+class EnsembleKeys:
+    """
+    Used internally by ensemble_learning.
+    """
+    # Serializing a trained model metadata.
+    MODEL = "model"
+    FEATURES = "feature_names"
+    TARGET = "target_name"
+    
+    # Classification keys
+    CLASSIFICATION_LABEL = "labels"
+    CLASSIFICATION_PROBABILITIES = "probabilities"
+
+
+class PyTorchInferenceKeys:
+    """Keys for the output dictionaries of PyTorchInferenceHandler."""
+    # For regression tasks
+    PREDICTIONS = "predictions"
+    
+    # For classification tasks
+    LABELS = "labels"
+    PROBABILITIES = "probabilities"
+    LABEL_NAMES = "label_names"
+
+
+class PytorchModelArchitectureKeys:
+    """Keys for saving and loading model architecture."""
+    MODEL = 'model_class'
+    CONFIG = "config"
+    SAVENAME = "architecture"
+
+
+class PytorchArtifactPathKeys:
+    """Keys for model artifact paths."""
+    FEATURES_PATH = "feature_names_path"
+    TARGETS_PATH = "target_names_path"
+    ARCHITECTURE_PATH = "model_architecture_path"
+    WEIGHTS_PATH = "model_weights_path"
+    SCALER_PATH = "scaler_path"
+
+
+class DatasetKeys:
+    """Keys for saving dataset artifacts. Also used by FeatureSchema"""
+    FEATURE_NAMES = "feature_names"
+    TARGET_NAMES = "target_names"
+    SCALER_PREFIX = "scaler_"
+    # Feature Schema
+    CONTINUOUS_NAMES = "continuous_feature_names"
+    CATEGORICAL_NAMES = "categorical_feature_names"
+
+
+class SHAPKeys:
+    """Keys for SHAP functions"""
+    FEATURE_COLUMN = "feature"
+    SHAP_VALUE_COLUMN = "mean_abs_shap_value"
+    SAVENAME = "shap_summary"
+
+
+class PyTorchCheckpointKeys:
+    """Keys for saving/loading a training checkpoint dictionary."""
+    MODEL_STATE = "model_state_dict"
+    OPTIMIZER_STATE = "optimizer_state_dict"
+    SCHEDULER_STATE = "scheduler_state_dict"
+    EPOCH = "epoch"
+    BEST_SCORE = "best_score"
+    HISTORY = "history"
+    CHECKPOINT_NAME = "PyModelCheckpoint"
+    # Finalized config
+    CLASSIFICATION_THRESHOLD = "classification_threshold"
+    CLASS_MAP = "class_map"
+    SEQUENCE_LENGTH = "sequence_length"
+    INITIAL_SEQUENCE = "initial_sequence"
+    TARGET_NAME = "target_name"
+    TARGET_NAMES = "target_names"
+
+
+class UtilityKeys:
+    """Keys used for utility modules"""
+    MODEL_PARAMS_FILE = "model_parameters"
+    TOTAL_PARAMS = "Total Parameters"
+    TRAINABLE_PARAMS = "Trainable Parameters"
+    PTH_FILE = "pth report "
+    MODEL_ARCHITECTURE_FILE = "model_architecture_summary"
+
+
+class VisionKeys:
+    """For vision ML metrics"""
+    SEGMENTATION_REPORT = "segmentation_report"
+    SEGMENTATION_HEATMAP = "segmentation_metrics_heatmap"
+    SEGMENTATION_CONFUSION_MATRIX = "segmentation_confusion_matrix"
+    # Object detection
+    OBJECT_DETECTION_REPORT = "object_detection_report"
+
+
+class VisionTransformRecipeKeys:
+    """Defines the key names for the transform recipe JSON file."""
+    TASK = "task"
+    PIPELINE = "pipeline"
+    NAME = "name"
+    KWARGS = "kwargs"
+    PRE_TRANSFORMS = "pre_transforms"
+    
+    RESIZE_SIZE = "resize_size"
+    CROP_SIZE = "crop_size"
+    MEAN = "mean"
+    STD = "std"
+
+
+class ObjectDetectionKeys:
+    """Used by the object detection dataset"""
+    BOXES = "boxes"
+    LABELS = "labels"
+
+
+class MLTaskKeys:
+    """Used by the Trainer and InferenceHandlers"""
+    REGRESSION = "regression"
+    MULTITARGET_REGRESSION = "multitarget regression"
+    
+    BINARY_CLASSIFICATION = "binary classification"
+    MULTICLASS_CLASSIFICATION = "multiclass classification"
+    MULTILABEL_BINARY_CLASSIFICATION = "multilabel binary classification"
+    
+    BINARY_IMAGE_CLASSIFICATION = "binary image classification"
+    MULTICLASS_IMAGE_CLASSIFICATION = "multiclass image classification"
+    
+    BINARY_SEGMENTATION = "binary segmentation"
+    MULTICLASS_SEGMENTATION = "multiclass segmentation"
+    
+    OBJECT_DETECTION = "object detection"
+    
+    SEQUENCE_SEQUENCE = "sequence-to-sequence"
+    SEQUENCE_VALUE = "sequence-to-value"
+    
+    ALL_BINARY_TASKS = [BINARY_CLASSIFICATION, MULTILABEL_BINARY_CLASSIFICATION, BINARY_IMAGE_CLASSIFICATION, BINARY_SEGMENTATION]
+
+
+class DragonTrainerKeys:
+    VALIDATION_METRICS_DIR = "Validation_Metrics"
+    TEST_METRICS_DIR = "Test_Metrics"
+
+
+class _OneHotOtherPlaceholder:
+    """Used internally by GUI_tools."""
+    OTHER_GUI = "OTHER"
+    OTHER_MODEL = "one hot OTHER placeholder"
+    OTHER_DICT = {OTHER_GUI: OTHER_MODEL}

ml_tools/_schema.py

@@ -2,7 +2,7 @@ from typing import NamedTuple, Tuple, Optional, Dict, Union
 from pathlib import Path
 
 from .custom_logger import save_list_strings
-from .keys import DatasetKeys
+from ._keys import DatasetKeys
 from ._logger import _LOGGER
 
 

ml_tools/custom_logger.py

@@ -1,6 +1,6 @@
 from pathlib import Path
 from datetime import datetime
-from typing import Union, List, Dict, Any
+from typing import Union, List, Dict, Any, Literal
 import traceback
 import json
 import csv
@@ -29,6 +29,8 @@ def custom_logger(
     ],
     save_directory: Union[str, Path],
     log_name: str,
+    add_timestamp: bool=True,
+    dict_as: Literal['auto', 'json', 'csv'] = 'auto',
 ) -> None:
     """
     Logs various data types to corresponding output formats:
@@ -36,10 +38,10 @@ def custom_logger(
     - list[Any]                    → .txt
         Each element is written on a new line.
 
-    - dict[str, list[Any]]        → .csv
+    - dict[str, list[Any]]        → .csv    (if dict_as='auto' or 'csv')
         Dictionary is treated as tabular data; keys become columns, values become rows.
 
-    - dict[str, scalar]           → .json
+    - dict[str, scalar]           → .json   (if dict_as='auto' or 'json')
         Dictionary is treated as structured data and serialized as JSON.
 
     - str                         → .log
@@ -49,29 +51,50 @@ def custom_logger(
         Full traceback is logged for debugging purposes.
 
     Args:
-        data: The data to be logged. Must be one of the supported types.
-        save_directory: Directory where the log will be saved. Created if it does not exist.
-        log_name: Base name for the log file. Timestamp will be appended automatically.
+        data (Any): The data to be logged. Must be one of the supported types.
+        save_directory (str | Path): Directory where the log will be saved. Created if it does not exist.
+        log_name (str): Base name for the log file.
+        add_timestamp (bool): Whether to add a timestamp to the filename.
+        dict_as ('auto'|'json'|'csv'): 
+            - 'auto': Guesses format (JSON or CSV) based on dictionary content.
+            - 'json': Forces .json format for any dictionary.
+            - 'csv': Forces .csv format. Will fail if dict values are not all lists.
 
     Raises:
         ValueError: If the data type is unsupported.
     """
     try:
+        if not isinstance(data, BaseException) and not data:
+            _LOGGER.warning("Empty data received. No log file will be saved.")
+            return
+        
         save_path = make_fullpath(save_directory, make=True)
         
-        timestamp = datetime.now().strftime(r"%Y%m%d_%H%M%S")
-        log_name = sanitize_filename(log_name)
+        sanitized_log_name = sanitize_filename(log_name)
         
-        base_path = save_path / f"{log_name}_{timestamp}"
-
+        if add_timestamp:
+            timestamp = datetime.now().strftime(r"%Y%m%d_%H%M%S")
+            base_path = save_path / f"{sanitized_log_name}_{timestamp}"
+        else:
+            base_path = save_path / sanitized_log_name
+        
+        # Router
         if isinstance(data, list):
             _log_list_to_txt(data, base_path.with_suffix(".txt"))
 
         elif isinstance(data, dict):
-            if all(isinstance(v, list) for v in data.values()):
-                _log_dict_to_csv(data, base_path.with_suffix(".csv"))
-            else:
+            if dict_as == 'json':
                 _log_dict_to_json(data, base_path.with_suffix(".json"))
+            
+            elif dict_as == 'csv':
+                # This will raise a ValueError if data is not all lists
+                _log_dict_to_csv(data, base_path.with_suffix(".csv"))
+            
+            else: # 'auto' mode
+                if all(isinstance(v, list) for v in data.values()):
+                    _log_dict_to_csv(data, base_path.with_suffix(".csv"))
+                else:
+                    _log_dict_to_json(data, base_path.with_suffix(".json"))
 
         elif isinstance(data, str):
             _log_string_to_log(data, base_path.with_suffix(".log"))
@@ -83,7 +106,7 @@ def custom_logger(
             _LOGGER.error("Unsupported data type. Must be list, dict, str, or BaseException.")
             raise ValueError()
 
-        _LOGGER.info(f"Log saved to: '{base_path}'")
+        _LOGGER.info(f"Log saved as: '{base_path.name}'")
 
     except Exception:
         _LOGGER.exception(f"Log not saved.")