Hi! Hello. I have a doubt. I have downloaded data from Chicago POI for 2021. To get a better sense of the data, I plotted businesses in Hyde Park, which I know well. Prior to doing this, I expected to see a lot of businesses in some streets (like the avenues; 53rd street in particular) and very few in other streets (those further away from avenues, particularly in the south). However, as you can see in the plot, businesses seem to be more or less evenly spread around Hyde Park, and usually, there is no more than one business per block. Am I doing something wrong when plotting the data?
Thanks in advance!
This topic was automatically generated from Slack. You can find the original thread here.
Hi @Martn_Serramo_Universidad_de_San_Andrs - looking at your account and I think I might know the reason you’re seeing this. Looks like you only downloaded Patterns data from the SafeGraph Shop. I’d recommend re-downloading Core Places as well and repeat this exercise with POIs in the Core Places download.
Keep in mind, that we have about 6.9 million POIs in Core Places and about 4.5 million POIs with Patterns data in the US. Not all POIs in Core Places will have Patterns. I’m guessing what you’ve done is mapped all the POIs near Hyde Park that were in the Patterns download, which is not representative of all the POIs that we would have for that area.
@Niki_Kaz Thanks, Niki! I So the idea with the Patterns data is not to generate a sample that replicates business density, but instead to evenly cover the map?
I apologize - I might have been a little unclear in my last comment. Core Places and Patterns are two very different datasets. Core Places provides business listing type information for POIs, like location name, address, lat/long, open/close hours, and so on.
Our Patterns dataset looks at foot traffic to these POIs. To my previous point, we do not have foot traffic data to all our POIs. That’s why you’ll see more POIs in our Core Places dataset compared to Patterns. Does that make more sense?