Combining open source, geodata and data visualization is one of our favorite exercise at Oslandia. The Pyris project totally corresponds to these criteria and allows us to play with INSEE (French national statistic institute) data.
Initialized by Damien Garaud, formerly data scientist at Oslandia (a big thanks to him!), this project continues to live at Oslandia through of our commitment to open source developments. Thanks to some recent works, Pyris aspect has dramatically changed.
Pyris is dealing with INSEE’s IRIS, which are homogeneous geographical areas that provide a basis for public census and every related socioeconomic analysis.
The technical stack used in Pyris is as follows:
- A PostGIS database that aims at storing INSEE socioeconomic data as well as IRIS geometries (the latter is coming from IGN).
- A Flask API that serves the data.
- A React.js front-end, which is deployed at data.oslandia.io/pyris as a demonstration website.
A Pyris instance is now listed within our data showroom, following the examples of Jitenshea and Deeposlandia.
The input data has been updated, the most recent data is now considered (2019 census and 2022 IRIS geometries). Furthermore the most important change is the front-end migration, from Bootstrap to React.js. The web application has been re-designed, in order to get a sexyer website. Here comes some before/after draws:
In the next months, we hope that these fruitful efforts will continue, so as to produce a more generic product. We plan to work on the following items in the project:
- simplify the data gathering pipeline;
- consider the Overseas (we currently only gets the metropolitan IRIS);
- consider several years of data so as to do temporal analysis;
- compare IRIS between each other or with respect to local/national averages;
- improve the software quality (lints, unit tests, CI).
Interested in Pyris? Do not hesitate to test it, open issues, submit Merge Requests or even reach us by email at email@example.com!