pydantic>2

[clausmichele]

Tests are passing, except one!

[clausmichele]

@christophreimer @ValentinaHutter we are using this version for out InterTwin use case, since it is necessary to have Pydantic > 2 in combination with HydroMT.

What do you think it would be the best approach to move towards? I know you have many other components relying on Pydantic 1.x, but we would like to find a way to publish the version based on Pydantic 2 somehow.

@iacopoff @jzvolensky

[ValentinaHutter]

Please make sure all tests are running successfully, otherwise we cannot merge the PR. We will then review it and create a new release.

Then we can start updating our internal repositories to go from pydantic<2 to pydantic>2.

[clausmichele]

@ValentinaHutter yes, there is one test failing, related to an old process graph from UC8 containing two experimental processes: fit_regr_random_forest and save_ml_model, That process graph is also not up to date, since the latest version (from UC8) is different and it's not using the process save_ml_model (which is the one creating issues with this test).

Here is the link to visualize the process graph used there:
https://editor.openeo.org/?server=https://openeo.eodc.eu/openeo/1.1.0&process=https://raw.githubusercontent.com/Open-EO/openeo-pg-parser-networkx/main/tests/data/graphs/fit_rf_pg_0.json&discover=1

Question: should we remove that process graph and update it with the latest used in UC8? Using that one the tests are passing. Let me know if you are still using the save_ml_model somewhere, if not I will commit the fix.

[ValentinaHutter]

@ValentinaHutter yes, there is one test failing, related to an old process graph from UC8 containing two experimental processes: fit_regr_random_forest and save_ml_model, That process graph is also not up to date, since the latest version (from UC8) is different and it's not using the process save_ml_model (which is the one creating issues with this test).

Here is the link to visualize the process graph used there: https://editor.openeo.org/?server=https://openeo.eodc.eu/openeo/1.1.0&process=https://raw.githubusercontent.com/Open-EO/openeo-pg-parser-networkx/main/tests/data/graphs/fit_rf_pg_0.json&discover=1

Question: should we remove that process graph and update it with the latest used in UC8? Using that one the tests are passing. Let me know if you are still using the save_ml_model somewhere, if not I will commit the fix.

Thanks for pointing this out! We replaced the save_ml_model process with using save_result(format= "GeoJSON") in order to make the pg more compliant to other process graphs. We still have the save_ml_model process available in the backend though - but I think we can remove it as the functionality is the same as in save_result.

For the tests, it should be fine to replace save_ml_model with save_result.

I will create a PR in our processes-spec to remove save_ml_model there.

[clausmichele]

@ValentinaHutter tests are passing, I replaced the process graph with the latest version used for UC8.