This analysis is of 24 product variables and 6 customer variables from over 1.37 million customers provided and collected by the H&M Group, with the goal of exploring what factors contribute to the whether or not a customer will purchase a particular article of clothing.

The provided data consists of three csv files: articles, customers, transactions_train.

• articles describes product information of the individual articles of clothing being sold.
• customers describes metadata pertaining to each of the recorded customers per each customer_id.
• transactions_train records all the transaction history.

I explore the following variables:

1. club_member_status: the loyalty rewards membership status
2. fashion_news_frequency: categorical variable of how frequent the customer receives fashion rate (NONE, Regularly, Other). Potentially a good indicator of interest in fashion.
3. age: self-reported discrete variable of customer age
4. postal_code: An encrypted categorical variable to represent geographical postal codes.
5. t_dat: date of transaction
6. price: transformed measurement of price of article purchased
7. product_type_no: the code for the particular article of clothing name such as trousers, dress, etc. This corresponds to the product_type_name
8. product_group_name: the portion of the body the article falls on (Garment Upper body, Garment Lower body, etc.)
9. graphical_appearance_no: the numerical value of the distribution of the pattern of the article which corresponds to graphical_appearance_name
10. colour_group_code: the numerical value of the color/pattern of the article which corresponds to colour_group_name

The data can be downloaded from Kaggle in the competition showcase, upon conclusion of the deadline: https://www.kaggle.com/competitions/h-and-m-personalized-fashion-recommendations/data

I also placed the data in this shared google drive folder: https://drive.google.com/drive/folders/1bwqk0qFqMTfvSHwN7Faa72evMaiikLBN?usp=sharing

ABOUT THE CODE:

final_visuals.RMD generates the final 3 plots for grading

timeseries.RMD explores various timeseries executed before selecting final 1

price_plots.RMD explores various plot combinations for the price visual before selecting final 1

age_price_plots.RMD explore the various plot combinations of the age and price scatterplot before selecting final 1

data_exploration.RMD was the initial exploration of the variables and data, including any general data engineering.