added numerical transformer

Author: RJ Agrawal

README.rst

@@ -11,7 +11,7 @@ In particular, it provides:
 
 1. A way to map ``DataFrame`` columns to transformations, which are later recombined into features.
 2. A compatibility shim for old ``scikit-learn`` versions to cross-validate a pipeline that takes a pandas ``DataFrame`` as input. This is only needed for ``scikit-learn<0.16.0`` (see `#11 <https://github.com/paulgb/sklearn-pandas/issues/11>`__ for details). It is deprecated and will likely be dropped in ``skearn-pandas==2.0``.
-3. A couple of special transformers that work well with pandas inputs: ``CategoricalImputer`` and ``FunctionTransformer``.
+3. A numerical transformer, ``NumericalTransformer``, that provides commonly used numerical transformation options. This helps serialize the ``DataFrameMapper``.
 
 Installation
 ------------
@@ -370,6 +370,29 @@ A ``DataFrameMapper`` will return a dense feature array by default. Setting ``sp
 The stacking of the sparse features is done without ever densifying them.
 
 
+Using Numerical Transformer
+****************************
+
+While you can use FunctionTransformation to generate artibtrary transformer but they cannot not serialized (pickled).
+NumericalTransformer takes function name as a string parameter and hence can be easily serialized.
+
+    >>> from sklearn_pandas import NumericalTransformer
+    >>> mapper5 = DataFrameMapper([
+    ...     ('children', NumericalTransformer('log')),
+    ... ])
+    >>> mapper5.fit_transform(data)
+    array([[1.38629436],
+               [1.79175947],
+               [1.09861229],
+               [1.09861229],
+               [0.69314718],
+               [1.09861229],
+               [1.60943791],
+               [1.38629436]])
+
+
+
+
 Changelog
 ---------
 

sklearn_pandas/__init__.py

@@ -3,3 +3,4 @@
 from .dataframe_mapper import DataFrameMapper  # NOQA
 from .cross_validation import cross_val_score, GridSearchCV, RandomizedSearchCV  # NOQA
 from .features_generator import gen_features  # NOQA
+from .transformers import NumericalTransformer # NOQA

sklearn_pandas/dataframe_mapper.py

@@ -166,6 +166,7 @@ def __setstate__(self, state):
         self.built_features = state.get('built_features', self.features)
         self.built_default = state.get('built_default', self.default)
         self.transformed_names_ = state.get('transformed_names_', [])
+        self.show_progressbar = state.get('show_progressbar', False)
 
     def _get_col_subset(self, X, cols, input_df=False):
         """

sklearn_pandas/transformers.py

@@ -1,6 +1,7 @@
 import numpy as np
 import pandas as pd
 
+from sklearn.base import BaseEstimator, TransformerMixin
 
 def _get_mask(X, value):
     """
@@ -12,3 +13,34 @@ def _get_mask(X, value):
         return pd.isnull(X)
     else:
         return X == value
+
+
+class NumericalTransformer(TransformerMixin):
+    """
+    Provides commonly used numerical transformers.
+    """
+    SUPPORTED_FUNCTIONS = ['log', 'log1p']
+
+    def __init__(self, func):
+        """
+        Params
+
+        func    function to apply to input columns. The function will be applied to each value.
+                Supported functions are defined in SUPPORTED_FUNCTIONS variable. Throws assertion
+                error if the not supported.
+        """
+        assert func in self.SUPPORTED_FUNCTIONS, \
+            f"Only following func arguments are supported: {self.SUPPORTED_FUNCTIONS}"
+        super(NumericalTransformer, self).__init__()
+        self.__func = func
+
+    def fit(self, X, y=None):
+        return self
+
+    def transform(self, X, y=None):
+        if self.__func == 'log1p':
+            return np.vectorize(np.log1p)(X)
+        elif self.__func == 'log':
+            return np.vectorize(np.log)(X)
+
+        raise ValueError(f"Invalid function name: {self.__func}")