US consumer prices data 2017-18 for sub-national CPI-U calculations using TiPRD model and R implementation

Published: 6 August 2021| Version 2 | DOI: 10.17632/w293jxpggj.2
Contributors:
,
,
,

Description

This data is a portion of a larger dataset, composed by over 120 million data points, collected by Starsift LLC for the Grocerybear Project (www.grocerybear.com) every day between January 2017 and May 2018 for over 50,000 unique items in about 750 commercial categories for eleven US cities: Boise, Honolulu, Houston, Las Vegas, Los Angeles, Orlando, Phoenix, Portland, San Francisco, Seattle, and Washington DC. This dataset is composed by 5 csv files, one for each CPI-U Entry Level Item disclosed: Apples, Bread, Butter, Cigarettes, and Coffee. Each file presents the following columns: Year, Month, Product name, Product code, City, Store Chain, Average price in the month. Store chains have been anonymized. This project also includes an R file to calculate sub-national consumer price indexes using the Time-interaction-Region Product Dummy (TiRPD) model.

Files

Steps to reproduce

Download the dataset, keeping the csv files in the "data" folder and run the R file. Requires R version 4.0.2 and installation of the following libraries: readr, dplyr, igraph, reshape2, ggplot2, grid, gridExtra.