Handling different NAICS years

[bl-young]

Stewi data tends to come in with unspecified NAICS year. In older versions we mapped all NAICS to 2012 NAICS with replace_naics_w_naics_from_another_year.

In the current recurisve approach, NAICS that don't align with the activity_schema (or 2012, when not specified as is the case with Stewi) get converted to blank in map_to_sectors() [here] (

flowsa/flowsa/flowby.py

Lines 1230 to 1239 in 74e9769

	for direction in ['ProducedBy', 'ConsumedBy']:
	fba_w_naics = (
	fba_w_naics
	.merge(naics_key,
	how='left',
	left_on=f'Activity{direction}',
	right_on='source_naics')
	.rename(columns={'target_naics': f'Sector{direction}'})
	.drop(columns='source_naics')
	)

)

Are there thoughts on how to best implement this moving forward?

[bl-young]

Went ahead and incorporated the old function 4f3e73d for now