Error when running cross_val_score - raise KeyError(key) from err KeyError: None

Question

I can run a manual fit/transform on the numeric pipeline without any errors:

fulldf = full_pipeline.fit_transform(train)

The error only comes when running"

scores = cross_val_score(full_pipeline_lr, X_train, y_train,cv = 2)

Full repex

import numpy as np
import pandas as pd
from sklearn.preprocessing import StandardScaler,PowerTransformer
from sklearn.pipeline import Pipeline, FeatureUnion
from sklearn.linear_model import LinearRegression
from sklearn.base import BaseEstimator, TransformerMixin

# load data
source = 'http://lib.stat.cmu.edu/datasets/boston_corrected.txt'
BostonHousing2 = pd.read_table(source, skiprows= 9)
boston = BostonHousing2.drop(['OBS.', 'MEDV'], axis=1)
boston.columns = map(str.lower, boston.columns)
boston["cmedv"] = np.log(boston["cmedv"])

# create stratified sample based on quantities of price
q = boston.cmedv.quantile([0, 0.25,0.5,0.75,1])
boston["cmedv_cat"] = pd.cut(boston["cmedv"], bins = q, right=True,
                             labels = [1,2,3,4], include_lowest = True)

from sklearn.model_selection import StratifiedShuffleSplit
traintest = StratifiedShuffleSplit(n_splits = 1, test_size=0.2, random_state=42)
for train_idx, test_idx in traintest.split(boston, boston["cmedv_cat"]):
    train= boston.loc[train_idx]
    test = boston.loc[test_idx]

train, test = train.drop(['cmedv_cat'], axis = 1), test.drop(['cmedv_cat'], axis = 1)
X_train = train.drop("cmedv", axis = 1)
y_train = train['cmedv']


# selects columns for pipeline
class ColumnSelect( BaseEstimator, TransformerMixin ):
    
    def __init__( self, feature_names):
        self._feature_names = feature_names
    
    def fit( self, X, y = None ):
        return self 
    
    def transform( self, X,  y = None ):
        return X[self._feature_names]

numerical_features =    ['lon', 'lat', 'crim', 'zn','lstat', 'age']

numerical_pipeline = Pipeline( steps = [
    ( 'num_select',  ColumnSelect(numerical_features) ),
    ( 'num_scale',   StandardScaler() ),
    ( 'yeo_johnson', PowerTransformer() )
    ])

full_pipeline = FeatureUnion( transformer_list = [ 
    ('numerical_pipeline',      numerical_pipeline),
    #('ordinal_pipeline',        ordinal_pipeline),
    #('categorical_pipeline',    categorical_pipeline)  
    ])

fulldf = full_pipeline.fit_transform(train)

# add linear regression
full_pipeline_lr = Pipeline( steps = [ 
    ( 'full_pipeline',  full_pipeline ),
    ( 'model',        LinearRegression() )   ])

# Fitting without cross validation works
full_pipeline_lr.fit( train, train['cmedv'] )  
y_pred = full_pipeline_lr.predict( test )
full_pipeline_lr.score( train, train['cmedv'] )  

# trying cross validation - getting error
from sklearn.model_selection import cross_val_score
scores = cross_val_score(full_pipeline_lr, X_train, y_train,cv = 2)

Error

Traceback (most recent call last):
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/sklearn/model_selection/_validation.py", line 531, in _fit_and_score
    estimator.fit(X_train, y_train, **fit_params)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/sklearn/pipeline.py", line 330, in fit
    Xt = self._fit(X, y, **fit_params_steps)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/sklearn/pipeline.py", line 296, in _fit
    **fit_params_steps[name])
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/joblib/memory.py", line 352, in __call__
    return self.func(*args, **kwargs)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/sklearn/pipeline.py", line 740, in _fit_transform_one
    res = transformer.fit_transform(X, y, **fit_params)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/sklearn/pipeline.py", line 953, in fit_transform
    results = self._parallel_func(X, y, fit_params, _fit_transform_one)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/sklearn/pipeline.py", line 983, in _parallel_func
    weight) in enumerate(transformers, 1))
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/joblib/parallel.py", line 1029, in __call__
    if self.dispatch_one_batch(iterator):
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/joblib/parallel.py", line 847, in dispatch_one_batch
    self._dispatch(tasks)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/joblib/parallel.py", line 765, in _dispatch
    job = self._backend.apply_async(batch, callback=cb)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/joblib/_parallel_backends.py", line 208, in apply_async
    result = ImmediateResult(func)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/joblib/_parallel_backends.py", line 572, in __init__
    self.results = batch()
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/joblib/parallel.py", line 253, in __call__
    for func, args, kwargs in self.items]
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/joblib/parallel.py", line 253, in <listcomp>
    for func, args, kwargs in self.items]
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/sklearn/pipeline.py", line 740, in _fit_transform_one
    res = transformer.fit_transform(X, y, **fit_params)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/sklearn/pipeline.py", line 367, in fit_transform
    Xt = self._fit(X, y, **fit_params_steps)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/sklearn/pipeline.py", line 296, in _fit
    **fit_params_steps[name])
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/joblib/memory.py", line 352, in __call__
    return self.func(*args, **kwargs)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/sklearn/pipeline.py", line 740, in _fit_transform_one
    res = transformer.fit_transform(X, y, **fit_params)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/sklearn/base.py", line 693, in fit_transform
    return self.fit(X, y, **fit_params).transform(X)
  File "<ipython-input-25-06c5d9af1bc4>", line 47, in transform
    return X[self._feature_names]
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/pandas/core/frame.py", line 2902, in __getitem__
    indexer = self.columns.get_loc(key)
  File "/Users/bxp151/opt/anaconda3/envs/boston/lib/python3.6/site-packages/pandas/core/indexes/base.py", line 2891, in get_loc
    raise KeyError(key) from err
KeyError: None

Answer 1

I modified:

self_.feature_names

to

self.feature_names

and it works. I'd love to know why that's the case if anyone knows.