abtestools

pyabtesting

An awesome README template to jumpstart your projects!
Explore the docs »

View Demo · Report Bug · Request Feature

Table of Contents

• About The Project
• Getting Started
  • Prerequisites
  • Installation
• Usage

About The Project

This library provides tools for AB Testing, very useful when working with marketing data

(back to top)

Getting Started

Prerequisites

To develop on this project, you will need to create a poetry environment with the needed dependencies

• install with Poetry

  poetry install
  

Installation

• git clone https://github.com/pablominue/pyabtesting.git
  

  or

  pip install abtestools
  

• Import main modules

  from abtestools import audience, test
  

(back to top)

Usage

import datetime

import pandas as pd

from abtestools.audiences import Audience
from abtestools.campaign import Campaign
from abtestools.test import Metric

data = pd.read_csv("tests/cookie_cats.txt", delimiter=",")

audience = Audience(
    users=data["userid"], group_mapping=dict(zip(data["userid"], data["version"]))
)

campaign = Campaign(
    audience=audience,
    metrics=[
        Metric(name="retention_1", type="discrete"),
        Metric(name="retention_7", type="discrete"),
    ],
    date_range=[
        datetime.datetime.today() - datetime.timedelta(days=x) for x in range(10)
    ],
)


def extract_data(date, metric_column: str, convert_bool: bool = True) -> dict:
    # Logic for each date calculation should be added here
    if convert_bool:
        data[metric_column] = data[metric_column].astype(int)
    return dict(zip(data["userid"], data[metric_column]))


for res in campaign.backfill(
    metric=Metric(name="retention_1", type="discrete"),
    extract_data=extract_data,
    metric_column="retention_1",
):
    print(res)