search

mattyschell / mobilelatlong2cloud

0 stars 0 forks source link

Are non-unique census block values correct? #4

Closed mattyschell closed 2 months ago

mattyschell commented 3 months ago

In the legacy application the returned values for census blocks are 4 digit block codes. Because census block codes are only unique within census tracts the 4-digit returned value for census blocks (alone) are not useful.

Here's a small selection of layer_name=CENSUSBLOCK feature_value = 2002.

image

For now we will assume that the consumers of this service want these non-unique values and always combine census blocks with other returned values like borough and census tract. The other possibility is that this is a mistake in the data and whoever last loaded the legacy application simply selected the wrong column to use as the feature_value source.

@mlipper please review and let me know if you think otherwise.

mlipper commented 2 months ago

Is there a proper spatial (e.g., "inner join" in non-spatial context) relationship between census block, tract, borough values returned by a successful point-in-polygon query? If so, then, for now at least, I think your assumption that the existing discrete set of data attributes is correct.

mattyschell commented 2 months ago

Yes, blocks nest within tracts and tracts nest within boroughs.

Great, thank you @mlipper closing this!