| | --- |
| | license: mit |
| | language: |
| | - en |
| | pretty_name: Consumer Price Index - United States (US) |
| | size_categories: |
| | - 1M<n<10M |
| | task_categories: |
| | - table-question-answering |
| | tags: |
| | - finance |
| | - text |
| | - parquet |
| | - time-series |
| | source_datasets: |
| | - https://download.bls.gov/pub/time.series/compressed/tape.format/ |
| | - bls.cpi.dateYYYYMM.gz |
| | --- |
| | # Changelog |
| | ### 2025-01-18 |
| | * I decided that I'll name the column survey, instead of consumer. I'll also set the value to the description, instead of the code. |
| | * I didn't realize that the pandas version of "Use this dataset" includes the filename. I'll remove the date from the filename. So that people do not have to change their code. |
| | * I have been updating this using the UI, but I will create a script in Spaces to update this from the BLS site. |
| |
|
| | ### 2025-01-12 |
| | While using the dataset, I discovered that BLS have two surveys combined into the CPI dataset. Hence, the new column consumer. I might decode this in the future. |
| | * CU - Urban Consumers |
| | * CW - Wage Earners |
