Tabular data is commonly represented in Python with numpy recarrays and pandas dataframes. FloPy recently added pandas as a core dependency. Work is underway to use pandas internally where expedient and support pandas dataframes across user-facing APIs where numpy recarrays were previously required.
Introduce a --pandas (short -P) pytest CLI option accepting values: yes, no, random. This controls the value of a boolean use_pandas fixture. Motivations:
allow parametrizing use_pandas argument to MFSimulation etc
allow controlling how package input data is provided in tests
Other miscellany:
some minor cleanup/reorg
use dest="KEEP" when configuring --keep option, access option from fixtures with request.config.option.KEEP
Tabular data is commonly represented in Python with numpy recarrays and pandas dataframes. FloPy recently added pandas as a core dependency. Work is underway to use pandas internally where expedient and support pandas dataframes across user-facing APIs where numpy recarrays were previously required.
Introduce a
--pandas(short-P) pytest CLI option accepting values:yes,no,random. This controls the value of a booleanuse_pandasfixture. Motivations:use_pandasargument toMFSimulationetcOther miscellany:
dest="KEEP"when configuring--keepoption, access option from fixtures withrequest.config.option.KEEP