search

DS4PS / cpp-529-spr-2020

Course shell for CPP 529 Data Practicum on Community Analytics for Spring 2020.
http://ds4ps.org/cpp-529-spr-2020
1 stars 1 forks source link

Transition Matrix #12

Open ecking opened 4 years ago

ecking commented 4 years ago

Hi there,

Can someone explain to me more about how to read the transition matrix?

I've googled it but I'm not quite understanding how it works in this instance.

image

AntJam-Howell commented 4 years ago

@ecking Column (1) contains counties that are in cluster group 1 in 2010. Row (1) contains counties contains counties that are predicted to be in cluster 1 based on 2000 data. The first element of 0.61 indicates that of the counties grouped into cluster 1 in 2010, 61% of those counties are also in the same cluster 1 in 2000. In other words, 61% of counties are in cluster 1 in 2000 and remain in cluster 1 10 years later in 2010. The high values on the diagonal (.61, .61, .85, and .67) suggest that the majority of counties remain in the same cluster grouping over the 10 year period.

Once you've correctly interpreted what each of the 4 cluster groups means, the sankey diagram will help you to see which counties are most likely to be experience gentrification (i.e. counties grouped into a low housing price cluster in 2000, which are then grouped into a high housing price cluster in 2010).