Kaggle Challenge: Predict Future Sales.

This dataset aims to publish the files that I will use on the Kaggle challenge called Predict Future Sales

Data

I have downloaded test and train data from the competition webpage.
I have downloaded shop and item information data from the English translations of @deargle from this post. Then I have made some changes in the data described in this R file.
I have collected historical USD/RUB rates from Investing.com. I have used the most recent data for the days which does not include a rate info (i.e. Saturdays and Sundays which markets are closed).
I have prepared a calendar depicting public holidays and weekends. Public Holiday info for Russia is collected from this site.

@kaggle.kazimanil_predict_future_sales_supplementary.calendar


CREATE TABLE calendar (
  "date" TIMESTAMP,
  "holiday" BIGINT,
  "weekend" BIGINT
);

@kaggle.kazimanil_predict_future_sales_supplementary.item_category


CREATE TABLE item_category (
  "item_id" BIGINT,
  "item_name_translated" VARCHAR,
  "item_cat1" VARCHAR,
  "item_cat2" VARCHAR
);

@kaggle.kazimanil_predict_future_sales_supplementary.shops_translated


CREATE TABLE shops_translated (
  "shop_id" BIGINT,
  "city" VARCHAR,
  "type" VARCHAR,
  "name" VARCHAR
);

@kaggle.kazimanil_predict_future_sales_supplementary.usd_rub


CREATE TABLE usd_rub (
  "date" TIMESTAMP,
  "cur_rate" DOUBLE
);