Skip to content

Data Storage

Mo Li edited this page Jul 15, 2021 · 5 revisions

useeior comes with built-in data and metadata stored in two places:

  • Data not manually created and not intended to be modified directly are generally stored in the data directory in R data archive format (.rda). For transparency and reproducibility, scripts are including in the data-raw as a record that were used to generate those files.

  • Crosswalks, indicators, field lists, and configuration files are stored in inst/extdata in text-based formats including comma-separated values (.csv) and YAML (.yml).

Satellite tables and indicators used in common USEEIO models included in the package release are stored externally on the EPA Data Commons, an Amazon AWS s3 server. Data files on the Data Commons are stored in parquet or csv format. Parquet is a storage format that is smaller and faster to read than traditional file types, such as a csv. Data files generated by other packages, such as flowsa or lciaformatter are made available on the Data Commons for use by useeior. If static model input files are not found locally, they will be downloaded from the Data Commons.

Clone this wiki locally