make sentence_transformer input optional dependency

Author: AnFreTh

mambular/preprocessing/prepro_utils.py

@@ -1,7 +1,6 @@
 import numpy as np
 import pandas as pd
 from sklearn.base import BaseEstimator, TransformerMixin
-from sentence_transformers import SentenceTransformer
 
 
 class CustomBinner(TransformerMixin):
@@ -228,20 +227,29 @@ def transform(self, X):
 class LanguageEmbeddingTransformer(TransformerMixin, BaseEstimator):
     """A transformer that encodes categorical text features into embeddings using a pre-trained language model."""
 
-    def __init__(self, model_name="paraphrase-MiniLM-L3-v2"):
+    def __init__(self, model_name="paraphrase-MiniLM-L3-v2", model=None):
         """
         Initializes the transformer with a language embedding model.
 
         Parameters:
-        - model_name (str): The name of the SentenceTransformer model to use.
+        - model_name (str): The name of the SentenceTransformer model to use (if model is None).
+        - model (object, optional): A preloaded SentenceTransformer model instance.
         """
         self.model_name = model_name
-        self.model = SentenceTransformer(model_name)
+        self.model = model  # Allow user to pass a preloaded model
+
+        if self.model is None:
+            try:
+                from sentence_transformers import SentenceTransformer
+
+                self.model = SentenceTransformer(model_name)
+            except ImportError:
+                raise ImportError(
+                    "sentence-transformers is not installed. Install it via `pip install sentence-transformers` or provide a preloaded model."
+                )
 
     def fit(self, X, y=None):
-        """
-        Fit method (not required for a transformer but included for compatibility).
-        """
+        """Fit method (not required for a transformer but included for compatibility)."""
         self.n_features_in_ = X.shape[1] if len(X.shape) > 1 else 1
         return self