For this challenge, each team will identify, extract, and convert United States data at the county level to .parquet that supports our temple estimation project.
Your discovered data must have a row for each county in the US with variables (columns) that your team feels would be of value to our project. You can use the county file provided in the repository as a guide.
- Fork this repository.
- Evaluate the other groups' described data they are building and make sure your team is not duplicating effort.
- Create a folder with a descriptive name of your data (E.g.,
residential_permits) - Create a data digestion script with the same name as your folder in the main section of the repository (E.g.,
residential_permits.py) - Write your
.parquetfiles into your respective folder with at least five chunked files, and no file is larger than 20 MB. - Create a data dictionary in the main section of the repository (E.g.,
residential_permits.mdthat describes each column in your data set. - Create at least two visualizations of your county data. One should be a map, and one should not be a map. Put the images in the
imagefolder. - Work with your team to have one pull request that returns this data to the central repository.
Your data must have the county fips ID column. It should not have the state and county name columns. It should have the same number of rows as the county_meta data.
| fips | permits_2020 | permits_2010 |
|---|---|---|
| 1045 | 45 | 38 |
| 1046 | 81 | 120 |
| 1081 | 123 | 19 |