-
Notifications
You must be signed in to change notification settings - Fork 13
pandahub developer meeting
Jan Ulffers edited this page Aug 4, 2022
·
7 revisions
04.08.2022
- updates regarding type casting? --> nothing new, yet
- Why does "convert_dataframes_to_dicts" try to encode all metadata as json? get_net_from_db_by_id fails, when metadata is not json --> This is due to recent changes - all metadata is now encoded to json. The convert function for old pandahub versions cannot fix this yet. The pandahub version before this change is released to master. The convert function will then be updated on develop.
30.06.2022
- data type casting in convert_dataframes_to_dicts: In the current implementation dataframe columns are always cast to their default type. This also means that, e.g. None gets cast to "None", when the column's dtype is str --> pandahub should not cast dtypes but only ensure bson conformity. Datatype checks need to be done in the REST-API.
- why are results not included when nets are written to the database but there are a tests that check that results are the same? --> The tests are used to compare power flow results for networks with different pandapower versions but the results are not written to the database
- MongoDB supports geospatial queries: For this reason geojson columns need to be converted to dicts when written to the database.