matthewgthomas / dplacer Goto Github PK

Starting with the next D-PLACE release, the data served by the web app will be fully imported from the D-PLACE/dplace-data repository, where it is curated.

Thus, rather than having to scrape the web, dplacer could access the data locally, via a download of a release or a clone of the repository (or a fork thereof). This would have several advantages:

• Other than the initial download/clone no internet access is required to work with dplacer.
• Less bandwidth and resource consumption of the D-PLACE app.
• People can add/correct data accessed via dplacer ...
• ... and if they work with a fork of the repository, they can even propose their changes to be fed into the main repository via pull requests.

The data in the repository is now batched into datasets, so to get data about all societies, you may have to visit the societies.csv files in several directories. But if this is inconvenient from R, you may also use the python API to navigate the data programmatically and put together a single CSV file which you can then feed to dplacer. We may also agree on a format for such a file and implement creation of such a file as subcommand of the dplace cli. Let me know what you think.