StevenReitsma

np.random.seed(25)
df_truck = generate_data(365, 300)
df_truck.head()

StevenReitsma

y = df_truck['l_100km']
X = df_truck[['area', 'headwind', 'weight_kg']]

X_train, y_train = map(lambda x: x[:265], [X, y])
X_test, y_test = map(lambda x: x[265:], [X, y])

StevenReitsma

from sklearn.base import BaseEstimator, TransformerMixin
from sklearn.pipeline import Pipeline, make_pipeline, FeatureUnion, make_union
from sklearn.ensemble import RandomForestRegressor

def get_categorical_glossary(x):
    n_unique = len(np.unique(x[x.notnull()]))
    codes = range(0, n_unique)
    return dict(zip(np.unique(x), codes))

class ColumnSelector(BaseEstimator, TransformerMixin):

StevenReitsma

# Numeric variables
numeric_pipeline = make_pipeline(ColumnSelector(['weight_kg']))

# Categorical variables
categorical_pipeline = make_pipeline(ColumnSelector(['area', 'headwind']), NumericEncoder())

# Combining numerical and categorical data preprocessing
preproc_union = make_union(numeric_pipeline, categorical_pipeline)

# Final pipeline

StevenReitsma

y_test_predictions = estimation_pipeline.predict(X_test)
prediction_errors = y_test - y_test_predictions

StevenReitsma

# Load packages
from rpy2.robjects import FloatVector
from rpy2.robjects.packages import importr

# Function to convert an R object to a Python dictionary
def robj_to_dict(robj):
    return dict(zip(robj.names, map(list, robj)))

StevenReitsma

cpm = importr('cpm')

StevenReitsma

cpm_result = cpm.detectChangePoint(FloatVector(prediction_errors), cpmType='Student', ARL0=1000)
cpm_result = robj_to_dict(cpm_result)

StevenReitsma

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline

StevenReitsma

train = pd.read_csv('train.csv', parse_dates=['Date'])
train.columns = train.columns.str.lower()
features = pd.read_csv('features.csv', parse_dates=['Date'])
features.columns = features.columns.str.lower()
features[['markdown1', 'markdown2', 'markdown3','markdown4', 'markdown5']] = features[['markdown1', 'markdown2', 'markdown3','markdown4', 'markdown5']].fillna(0)
stores = pd.read_csv('stores.csv')
stores.columns = stores.columns.str.lower()
df_base = (
    train
    .merge(features, how='inner')