gmreincglm/usta-feeds-samples
Dated samples of United States public-record change files 15 samples, one folder per family. Each folder holds sample.csv, sample.json and a README naming the source, the columns, the sealing date and the row count. Every file is a change file, not a snapshot. We seal dated copies of a public source, compare two copies, and keep what appeared, what stopped being listed, and what quietly changed in between. Most of these sources publish only the list as it stands today and… See the full description on the dataset page: https://huggingface.co/datasets/gmreincglm/usta-feeds-samples.
Dated samples of United States public-record change files
15 samples, one folder per family. Each folder holds sample.csv, sample.json and a README naming the source, the columns, the sealing date and the row count.
Every file is a change file, not a snapshot. We seal dated copies of a public source, compare two copies, and keep what appeared, what stopped being listed, and what quietly changed in between. Most of these sources publish only the list as it stands today and overwrite the last one, so the history is not recoverable once it is gone.
Row counts above are data rows. Four files carry extra lines that a source publisher requires us to keep with the rows; those lines are not data and each folder's README says which.
These samples exist to test whether the file loads in your pipeline. They are too small to answer a research question, and they do not establish customer activity or business performance.
Sources, coverage dates, column lists, delivery terms and prices for the complete files are documented per family at https://ustechautomations.com/feeds/
Licence: See LICENSE in this repository.
Questions: admin@ustechautomations.com
Tooling: CSV comparison and coverage check · Run and reproduce