# Forecasting_for_Action — Data

来源：https://www.kaggle.com/competitions/forecasting-for-action/data

采集时间：2026-10-19T17:43:37Z

For this challenge, you will be predicting multiple years worth of sales for various Kaggle-branded stickers from different fictitious stores in different (real!) countries. This dataset is completely synthetic, but contains many effects you see in real-world data, e.g., weekend and holiday effect, seasonality, etc.

Good luck!

## Files

*   **train.csv** - the training set, which includes the sales data for each date-country-store-item combination. 
*   **test.csv** - the test set; your task is to predict the corresponding item sales (`num_sold`) for each date-country-store-item combination. Note the Public leaderboard is scored on the first year of the data, and the Private on the remaining.
*   **sample_submission.csv** - a sample submission file in the correct format