Grocery Sales Forecasting

This repo is the work summary of Corporación Favorita Grocery Sales Forecasting.

Report

Under report folder, contain source code for reporting.

• report/README.Rmd: R markdown to provide data insight of Corporación Favorita Grocery Sales data.

Further report detail under report folder.

Data Processing

Under process folder, contain source code for data processing.

• model/data_integration.R: R Script to integrated train.csv and test.csv with stores.csv, items.csv and transactions.csv.

Further processing detail under process folder.

Data Modeling

Under model folder, contain source code for data modeling.

Further modeling detail under model folder.

Data Source

Data source from Corporación Favorita, access the data from this competition Corporación Favorita Grocery Sales Forecasting in Kaggle.

That is 7 data files:

• train.csv: Training data, includes the target unit_sales by date, store_nbr, and item_nbr and a unique id to label rows, onpromotion column tells whether that item_nbr was on promotion for a specified date and store_nbr.

• test.csv: Test data, with the date, store_nbr, item_nbr combinations that are to be predicted, along with the onpromotion information.

• stores.csv: Store metadata, including city, state, type and cluster(grouping of similar stores).

• items.csv: Item metadata, including family, class, and perishable(have a score weight of 1.25; otherwise, the weight is 1.0).

• transactions.csv: The count of sales transactions for each date, store_nbr combination. Only included for the training data timeframe.

• oil.csv: Daily oil price, includes values during both the train and test data timeframe.

• holidays_events.csv: Holidays and Events, with metadata.