I had used SafeGraph's shapefile before to easily match to other GIS files. Do you know if it is possible to get these?

Help
,
#1

@Ryan_Fox_Squire_SafeGraph I had used SafeGraph’s shapefile before to easily match to other GIS files. Do you know if it is possible to get these?

#2

@Luis_E_Quintero what data specifically do you want as a shapefile? we can possibly make it happen.

do you already have an account on shop.safegraph.com? what email is your account under?

#3

Can you merge the ids with the census tiger block groups? I haven’t tried but this is my plan.

#4

Hi @Jude_Bayham_Colorado_State_U and @Luis_E_Quintero

2 things I can add here

  1. SafeGraph has all of the census block group geojson files available for your use if you want them. I just added a new row “Open Census Data”
    Open Census Data: Everything you need to know to get started

  2. This notebook shows in python how to join individual POI to the census block group data (which also allows you to match counties and census tracts).
    Google Colab

If you know some python (or can stumble through it) and If you are doing this on a small scale you can clone that notebook on GoogleCoLab and run it directly on their servers without installing anything yourself. (

#5

So, census block groups aren’t already appended to the data, but you can easily do this using the Open Census Data CBG geojson and the lat/long coordiantes from Core Places

#6

I was thinking a shapefile with POI polygons and related visits by day and duration of visits numbers in the attributes. This would be useful to quickly make maps that show the evolution over time of traffic in specific points

#7

And to join it with other county level data (I do not know python and joining using shapefiles with QGIS has worked for me so far). However, if county indicators for each POI were in the data that would be also useful and I wouldn’t have to do the joins by lat and long (in the data there is a city indicator but not county… Am I missing anything?)

#8

FWIW…
I recently tried using the WKT geometry representation included in the most recent Open Census data release and found it a hassle to deal with. The US Census now offers a concise set of “cartographic boundary files” that may be sufficient for most analyses at the block group (CBG) and larger. I found these much easier to work with for spatial joins & summaries.

In particular, the 1 : 500,000 (national) geodatabase offers a nice tradeoff between size and completeness, and contains most layers of interest. This can be loaded straight into QGIS or Postgresql (via org2ogr), and contains the required join columns (i.e., state, county, and CBG fips):
https://www2.census.gov/geo/tiger/GENZ2020/gdb/cb_2020_us_all_500k.zip [245 MB]

1 Like
#9

Thanks for sharing this @Christian_Gunning_University_of_Georgia. We’re also working on adding additional publicly available datasets to Dewey so we’ll look into this as well.