Script to quickly process and load into memory anonymous sales conversion data from Kaggle

preprocessing_sales_conversion_data.py

import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
import requests
from io import StringIO

COLUMNS_TO_DROP = ["ad_id", "xyz_campaign_id", "fb_campaign_id", "Total_Conversion"] # drop to avoid potential data leakage

TARGET = ["Approved_Conversion"]
# a hosted version of sales optimization conversion data from an anonymous organization, originally from Kaggle
# at https://www.kaggle.com/loveall/clicks-conversion-tracking
data_url = "https://raw.githubusercontent.com/ychennay/ychennay.github.io/master/KAG_conversion_data.csv"
# read data into memory and drop columns
data_string = requests.get(data_url).content
conversions_df = pd.read_csv(StringIO(data_string.decode("utf-8"))).drop(columns=COLUMNS_TO_DROP)
conversions_df["bias"] = 1 # add a bias/intercept column
# define the target and features
y = conversions_df[TARGET]
X = conversions_df.loc[:, ~conversions_df.columns.isin(TARGET)]

# using dictionary convert columns into categorical data types
convert_dict = {'gender': "category", 'interest':"category", "age": "category"}
# get dummy features for categorical variables - X is 1143 x 47, y is 1143 x 1
X = pd.get_dummies(X.astype(convert_dict), drop_first=True)