Example of usage model from sklift.models in sklearn.pipeline


This is a simple example on how to use sklift.models with sklearn.pipeline.

The data is taken from MineThatData E-Mail Analytics And Data Mining Challenge dataset by Kevin Hillstrom.

This dataset contains 64,000 customers who last purchased within twelve months. The customers were involved in an e-mail test:

  • 1/3 were randomly chosen to receive an e-mail campaign featuring Mens merchandise.
  • 1/3 were randomly chosen to receive an e-mail campaign featuring Womens merchandise.
  • 1/3 were randomly chosen to not receive an e-mail campaign.

During a period of two weeks following the e-mail campaign, results were tracked. The task is to tell the world if the Mens or Womens e-mail campaign was successful.

The full description of the dataset can be found at the link.

Firstly, install the necessary libraries:

In [1]:
!pip install scikit-uplift xgboost==1.0.2 category_encoders==2.1.0 -U

For simplicity of the example, we will leave only two user segments:

  • those who were sent an e-mail advertising campaign with women's products;
  • those who were not sent out the ad campaign.

We will use the visit variable as the target variable.

In [2]:
import pandas as pd
from sklift.datasets import fetch_hillstrom

%matplotlib inline

bunch = fetch_hillstrom(target_col='visit')

dataset, target, treatment = bunch['data'], bunch['target'], bunch['treatment']

print(f'Shape of the dataset before processing: {dataset.shape}')

# Selecting two segments
dataset = dataset[treatment!='Mens E-Mail']
target = target[treatment!='Mens E-Mail']
treatment = treatment[treatment!='Mens E-Mail'].map({
    'Womens E-Mail': 1,
    'No E-Mail': 0

print(f'Shape of the dataset after processing: {dataset.shape}')
Shape of the dataset before processing: (64000, 8)
Shape of the dataset after processing: (42693, 8)
recency history_segment history mens womens zip_code newbie channel
0 10 2) $100 - $200 142.44 1 0 Surburban 0 Phone
1 6 3) $200 - $350 329.08 1 1 Rural 1 Web
2 7 2) $100 - $200 180.65 0 1 Surburban 1 Web
4 2 1) $0 - $100 45.34 1 0 Urban 0 Web
5 6 2) $100 - $200 134.83 0 1 Surburban 0 Phone

Divide all the data into a training and validation sample:

In [3]:
from sklearn.model_selection import train_test_split

X_tr, X_val, y_tr, y_val, treat_tr, treat_val = train_test_split(
    dataset, target, treatment, test_size=0.5, random_state=42

Select categorical features:

In [4]:
cat_cols = X_tr.select_dtypes(include='object').columns.tolist()
['history_segment', 'zip_code', 'channel']

Create the necessary objects and combining them into a pipieline:

In [5]:
from sklearn.pipeline import Pipeline
from category_encoders import CatBoostEncoder
from sklift.models import ClassTransformation
from xgboost import XGBClassifier

encoder = CatBoostEncoder(cols=cat_cols)
estimator = XGBClassifier(max_depth=2, random_state=42)
ct = ClassTransformation(estimator=estimator)

my_pipeline = Pipeline([
    ('encoder', encoder),
    ('model', ct)

Train pipeline as usual, but adding the treatment column in the step model as a parameter `model__treatment`.

In [6]:
my_pipeline = my_pipeline.fit(
/Users/Maksim/Library/Python/3.6/lib/python/site-packages/sklearn/pipeline.py:354: UserWarning: It is recommended to use this approach on treatment balanced data. Current sample size is unbalanced.
  self._final_estimator.fit(Xt, y, **fit_params)

Predict the uplift and calculate the [email protected]%

In [7]:
from sklift.metrics import uplift_at_k

uplift_predictions = my_pipeline.predict(X_val)

uplift_30 = uplift_at_k(y_val, uplift_predictions, treat_val, strategy='overall')
print(f'[email protected]%: {uplift_30:.4f}')