GeoID miss when joining with census blocks?

enigma-io / smoke-signals-model

The Machine Learning Algorithms that power Smoke Signals.

http://labs.enigma.io/smoke-signals

38 stars 16 forks source link

GeoID miss when joining with census blocks? #3

Closed stuartlynn closed 9 years ago

stuartlynn commented 9 years ago

Trying to map the results of the analysis here by joining the smoke

It looks like perhaps there is a miss match between the GeoID's in the smoke risk scores and the census blocks which is specific to certain states.

https://team.cartodb.com/u/stuartlynn/viz/7ba9f95e-5b04-11e5-8fc5-0e0c41326911/map

abelsonlive commented 9 years ago

Thanks for reaching out! What census shapefile are you using?

stuartlynn commented 9 years ago

Its the tl_2014_block_groups dataset in CartoDB.

I think I have figured it out though. It looks like the block groups in that file are padded so 11170306092 in the smoke-alarm-risk data should be 011170306092 in the census data. Should be able to resolve here now.

Thanks for the quick reply.

stuartlynn commented 9 years ago

That seems to have fixed it

http://cdb.io/1iqB82a