Cluster analysis - base on commute time

Question

Hi All,

I have a simple analysis: find a group/cluster of offices within 30mins(<-flex, might change) each other.

Given a group of offices (1 file) within several different state the audience wants to know if they can see a cluster/group base on drive time within 30mins of each other offices.

This is what I have done so far:

1. Get the SpatialObject for each office

2. append the offices to each other to get the pairing

3. filter only pair within the same state( there might be offices cross state that might fall within 30min criteria but trying to make it simple here for now)

4. use the distance tool to calculate driving distance within 30mins of each other

Base on this result audience would need to filter on every offices to see the possible pairing.

Maybe I'm overthinking this but are there further steps I can perform to make the analysis easier to digest for the audiences? Like group offices into clusters? Any suggestions would be great, thanks.

AnalyticJack · Answer

Hey BosKev,

Even though this thread is a little old, I thought I would reply as I've recently joined a company called TravelTime that offers a Plugin directly to Alteryx that offers exactly what you are looking for!

Please take a look at www.traveltime.com for further info.

Cheers,

Analytic Jack

PhilipL · Answer

Try using Find Nearest and feed your input file (with point locations) to both the Target and Universe inputs. Set the max distance to 30 Drive Distance miles, and bump the number "nearest points to find" up to as high as you want (the higher the number, the longer to process). Be sure to ignore 0 distance matches.

This will not do your within state filter, but you can always do that after, or just leave it as is to be accurate. In non-COVID times, crossing state lines isn't usually a big deal. 🙂

wlau · Answer

The steps you listed sounds fine to me.   I can see your point about having to filter one office at a time.

Have you thought about using a batch macro to perform the steps for each office?  It can combine all your results into one output.

In terms out nice output that easy to digest, I suggest using a map. So you'll need to include the spatial objects in the output, that is, the centroid for each office and its 30min drive time contour shape. This will open up some more options for you in terms of presentation.  You can try plotting these inside an interactive map in Alteryx (or just use a Browse which is really simple).  Or export the data to files and import to another tool that supports maps.

Hope this helps.