Ensure API response can handle matching datasets lacking pipeline data

[alyssadai]

We need to ensure the presence of NaNs for pipeline_name and pipeline_version don't break things:

api/app/api/crud.py

Lines 202 to 220 in 9913736

	pipeline_grouped_data = (
	group.groupby(
	[
	"sub_id",
	"session_id",
	"session_type",
	"pipeline_name",
	],
	dropna=True,
	)
	.agg(
	{
	"pipeline_version": lambda x: list(
	x.dropna().unique()
	)
	}
	)
	.reset_index()
	)