Our Meetup Metadata - Member location

Public Relations
#1

Here’s a list of nearby cities and their respective counts of the number of our Meetup’s members living there.

|Dallas, TX|1360|
|Carrollton, TX|497|
|Plano, TX|377|
|Irving, TX|309|
|Richardson, TX|156|
|Denton, TX|141|
|Fort Worth, TX|141|
|Frisco, TX|134|
|Lewisville, TX|129|
|Arlington, TX|123|
|Addison, TX|114|
|McKinney, TX|107|
|Coppell, TX|91|
|Allen, TX|73|
|The Colony, TX|62|
|Flower Mound, TX|60|
|Garland, TX|58|
|Grapevine, TX|50|
|Keller, TX|45|
|Euless, TX|40|
|Little Elm, TX|34|
|Grand Prairie, TX|34|
|Southlake, TX|30|
|Wylie, TX|27|
|Mesquite, TX|27|
|Bedford, TX|26|
|Rowlett, TX|24|
|Hurst, TX|21|
|Roanoke, TX|21|
|North Richland Hills, TX|15|
|Aubrey, TX|13|
|Duncanville, TX|12|
|Prosper, TX|11|
|Sachse, TX|11|
|Colleyville, TX|11|
|Forney, TX|10|
|Tyler, TX|9|
|Mansfield, TX|8|
|Argyle, TX|8|
|Rockwall, TX|8|
|Justin, TX|7|
|Waxahachie, TX|6|
|Melissa, TX|6|
|Lake Dallas, TX|6|
|Greenville, TX|6|
|Sherman, TX|5|
|Cedar Hill, TX|4|
|Sunnyvale, TX|4|
|Kaufman, TX|4|
|Lancaster, TX|4|
|Terrell, TX|4|
|Cleburne, TX|4|
|Burleson, TX|4|
|Gainesville, TX|4|
|Red Oak, TX|4|
|Midlothian, TX|4|
|Waco, TX|3|
|Weatherford, TX|3|
|Desoto, TX|3|
|Valley View, TX|3|
|Royse City, TX|3|

#2

Is it possible to get this data broken down by zip code?

1 Like
#3

Please. There are free-to-use GIS systems out there which would plot this info if provided by ZIP.

#4

Can’t do zip codes, but can do the longitudes and latitudes associated with each user which is more specific to each user than the city data.

#5

Was trying to post it all here but I’m 90,000 characters over the limit. Let me know if there’s a great community spot where I can put a spreadsheet or DM me an email address and I’ll send it over.

Here’s example:
33,-96.78,Dallas
32.87,-96.98,Irving
32.89,-96.86,Dallas
33.17,-97.11,Denton
32.83,-96.77,Dallas
32.83,-96.77,Dallas
32.96,-96.98,Coppell
32.82,-96.75,Dallas

#6

Gimme a couple days and I’ll see about correlating each lon/lat with a zip code and then getting the counts.

#7

DONT SHARE PII

From a Personally Identifiable Information perspective, I would recommend rounding off the lat,lon info or adding a random fudge factor (I recommend adding +RANDBETWEEN(-100,100)/10000 to both the lat and lon) to obscure the true location.

That fudge factor would take a single location and obscure it to somewhere within an area about the size of a neighborhood. Here is the same lat/lon location (33,-97) randomized 100 times using that fudge factor and uploaded to GPS Visualizer. Accurate enough for determining which part of the Metroplex, but no PII released.

image
image.jpg893×834 739 KB

WHERE TO SHARE THE DATA

Once that’s done, you could upload the fudged data (but not the original location) to a Google Sheet on drive.google.com and share the URL as read only.

3 Likes
#8

This website requires a Google API key and I’m seeing a charge of $5 for every 1,000 requests. Might you happen to already have a key and would be so generous?

Do you choropleth? with zip codes?
#9

What, exactly, is the purpose of this member scatterplot? I agree that PII data is beginning to become concerning. Not yet…but the drill downs likely need to be kept to city level.

#10

Here’s a heatmap of the geocoordinates. Everybody getting the same interpretation about freeway boundaries that I am? The concentration around downtown Dallas is probably an artifact created by Meetup or where people work. I need to test how/if a person’s geocoordinates change with Meetup usage.

image
image.png900×640 552 KB

1 Like
#11

RE: PII

From https://dallasmakerspace.org/wiki/Rules_and_Policies#Privacy_Policy

In compliance with law and our bylaws, DMS will provide Member’s names and physical adress only, in response to a valid request for member information.

Ref: https://dallasmakerspace.org/wiki/Bylaws#Article_8_-_Corporate_Books.2C_Records_and_Reports

#12

I think I now have a really good handle on making heat maps using the Google Maps API and won’t be able to produce better representations of this data than this. But if you do know how to improve or would like to see more representations like colored zip code maps (still in the works at a casual pace) please let me know as I would like to continue to make this work better.

image
image.png1304×778 812 KB



Zoomed in slightly
image
image.png1430×838 1.09 MB

6 Likes
#13

Sanity check:

1 Like
#14

Dear God, WTH happened to Fort Worth? Are they ok? Was FEMA notified?
Only razzin’ back amigo. :grin:

What about…
Higher income vs. lower income areas?
Containerization within the main freeways?
Grand Prarie (23 min, 194K) vs. Plano (22 min, 286K): what’s that about?
Heat pockets in Denton and Frisco jumping out at me… perhaps there are niche communities of artists or techies our hard-working, outgoing PR team members will want to focus their energies on.

I don’t know at the outset what exactly these types of analyses will show, but I think it’s worth us all taking a look and shouting out if you spot something. I think we’ll be a lot better off understanding who comprises our organization than not.

If you have questions, I’d like to drill down deeper into the data to answer them.

#15

Can the data be normalized for population (as a function of geographic location)? I would think this would be a logical approach to resolving the issue highlighted in the xkcd comic strip.

https://www.texas-demographics.com/counties_by_population

#16

https://www.opendatanetwork.com/entity/1600000US4819000/Dallas_TX/geographic.population.density?year=2017

#17

Great minds. I think counties are too big and want to do it by zip codes.

#18

@HankCowdog @darrent @Draco
Would you all say you agree or disagree with my interpretations of different income areas, freeway containerization, comparison of Grand Prairie vs. Plano, & Denton, Frisco, (and West Plano too) being hot spots?

Would really appreciate the feedback to know if I’m interpreting the data correctly or not.

1 Like
#19

Eric, short of having those other datasets (ie. population, income, distance) tied to the Meetup data, interpretation most assuredly will be guess work on any ones part.

I think the interpretation about containerization is very likely to be true, but data on distance (miles and time) would really be needed to confirm. This data can likely be had on a zipcode by zipcode basis.

Your note on GP (34 meetup members) and P (377 meetup members) is certainly interesting …

“Grand Prarie (23 min, 194K) vs. Plano (22 min, 286K): what’s that about?”

GP has 2/3 the population of P, but GP has only 10% the number of meetup members as P. But both are roughly the same distance from the Makerspace. Now if the income is included … GP ($55k), P ($103k), does that explain it? I don’t know … For both locations the ‘best route’ according to google maps includes a toll fee. That might be more important to those in GP.

Worth noting, the meetup population density in Plano is skewed to the west of 75 and the distance on google maps is from the center of town (east side of 75). Thus for the example above, the distance / time traveled is likely shorter for most of the Plano contingent in the Meetup dataset. I presume this is one of the reasons you are interested in getting the data on a zipcode basis.

As with most of what I say and write, I’m fairly certain it’s just a bunch of bs.

2 Likes
#20

Nah dude, totally brilliant and insightful. Thank you!

It’s really easy to get the actual drive time distances for each person through Google’s API. I’ll check out if there’s ways to include toll options too when playing around with that.

It’ll be super easy to normalize for population and income too. Just need to crack this zip code issue in my spare time unless anybody knows a better way.