StevenReitsma

import warnings
warnings.filterwarnings('ignore')
predictions, test_indices = cross_val_predict(multi, X_pd_num, y=y, cv=multiple_ts_split, n_jobs=-1)

StevenReitsma

pred_lgb = pd.DataFrame(predictions)
pred_lgb.columns = ['pred_lgb1', 'pred_lgb2']
df_pred_fc = df_features.copy().reset_index(drop=True)
for col in pred_lgb.columns:
    df_pred_fc[col] = np.nan
    df_pred_fc.loc[test_indices, col] = pred_lgb[col].values
validate = df_pred_fc.dropna()
selected = validate[['store', 'dept']].drop_duplicates().sample(1)
(
    validate

StevenReitsma

(
    validate
    .merge(selected, how='inner')
    .set_index('date')
    [['pred_lgb2', 'weekly_sales2']]
    .plot(figsize=(10,4))
)
plt.title(selected.reset_index(drop=True))

StevenReitsma

from sklearn.model_selection._split import check_cv
from sklearn.model_selection._validation import _fit_and_predict
from sklearn.externals.joblib import Parallel, delayed, logger
from sklearn.preprocessing import LabelEncoder
from sklearn.base import is_classifier, clone
from sklearn.utils import indexable
import scipy.sparse as sp
def cross_val_predict(estimator, X, y=None, groups=None, cv=None, n_jobs=1,
                      verbose=0, fit_params=None, pre_dispatch='2*n_jobs',
                      method='predict'):

StevenReitsma

import pandas as pd
import numpy as np
from sklearn.pipeline import Pipeline, FeatureUnion
from sklearn.preprocessing import OneHotEncoder, StandardScaler
transformer = Pipeline([
    ('features', FeatureUnion(n_jobs=1, transformer_list=[
        # Part 1
        ('boolean', Pipeline([
            ('selector', TypeSelector('bool')),
        ])),  # booleans close

StevenReitsma

# Example dataframe with different data types.
df = pd.DataFrame({
    'boolean_column': [True,False,True,False], 
    'integer_column': [1,2,3,4],
    'float_column': [1.,2.,3.,4.]
})
# Selecting booleans
boolean_columns = df.select_dtypes(include=['bool'])
# Selecting numericals
numerical_columns = df.select_dtypes(include=[np.number])

StevenReitsma

from sklearn.base import BaseEstimator, TransformerMixin
class TypeSelector(BaseEstimator, TransformerMixin):
    def __init__(self, dtype):
        self.dtype = dtype
    def fit(self, X, y=None):
        return self
    def transform(self, X):
        assert isinstance(X, pd.DataFrame)
        return X.select_dtypes(include=[self.dtype])

StevenReitsma

# Transforming a column into a category type
df = pd.DataFrame({'eye_color': ['green', np.nan, 'blue', 'brown']})
df['eye_color'] = df['eye_color'].astype('category')

StevenReitsma

# Inspecting the category codes.
df['eye_color'].cat.codes
# Inspecting the categories
df['eye_color'].cat.categories

StevenReitsma

class StringIndexer(BaseEstimator, TransformerMixin):
    def fit(self, X, y=None):
        return self
    def transform(self, X):
        assert isinstance(X, pd.DataFrame)
        return X.apply(lambda s: s.cat.codes.replace(
            {-1: len(s.cat.categories)}
        ))