Are we returning pandas dfs across the board now?

Are we returning pandas dfs across the board now?

We are returning pydantic models, which usually contain a dataframe that is validated with pandera.

I might have appreciated a datetime format rather than string dates (datetime.datetime.strptime(data.period["Tidsperiod (fr.o.m)"][0],"%Y-%m-%d %H:%M:%S"), but we can leave that parsing for the parent class, if needed.

Could you point to where in the code? If it is validtime from say Mesan where get_multipoint expects a string parameter, I don't think it's a good idea. Do you mean the index in the dataframes? Isn't that a date index already?

I mean in the data.period from and to fields :)

Let's think about it, as Strang has those fields too.