Adjusted method ordinal_encoding when it get's a dataframe with dtype='string' (#277)

The problem occurs because the dataframe contains `dtype='string'` parameter. With this, the `unique()` method returns `StringArray` object, no np array.
Instead of using `tolist` method, it was included the `list` call outside.

Co-authored-by: Leonardo <leonardo.fiedler@senior.com.br>

Author: leonardoFiedler

category_encoders/ordinal.py

@@ -319,8 +319,8 @@ def ordinal_encoding(X_in, mapping=None, cols=None, handle_unknown='value', hand
             for col in cols:
 
                 nan_identity = np.nan
-
-                categories = X[col].unique().tolist()
+                
+                categories = list(X[col].unique())
                 if util.is_category(X[col].dtype):
                     # Avoid using pandas category dtype meta-data if possible, see #235, #238.
                     if X[col].dtype.ordered: