7,613
edits
Changes
Jump to navigation
Jump to search
Urban Start-up Agglomeration and Venture Capital Investment (view source)
Revision as of 18:18, 9 August 2019
, 18:18, 9 August 2019→New Work
Note that everything uses Geographies (except to find centroids), and pair lengths and hull areas are scaled so that they are in hm and hectares (hm2) [https://en.wikipedia.org/wiki/Hectometre], as this is close to being a city block length and square block [https://en.wikipedia.org/wiki/City_block#Grid_plan] at least in Houston, TX (note that a block in Chicago is 2 Houston blocks and there is no standard block definition).
Then load up leveldefinitions.txt. Note that we are using more levels than before, with finer grained levels at the bottom end:
Level Label Target
1 100msq 0.01
2 1blfront 0.1
3 1blsq 1
4 5blsq 5
5 10blsq 10
6 25blsq 25
7 50sqbl 50
8 1kmsq 100
9 5kmsq 400
10 10kmsq 1000
11 20min rule 35000
Then build the nohulls as level 0 and the allhull as level 12. Note that we are going to have to throw out panels or observations with too few points per city-year later, as these can have singletons, multitons, or pairs as their allhulls. This can be done with nohull !=0. Also nohulls can be built using hcllayer then it will contain pairs, or from colevelforcircles. I opted for the later, so that it only contains singletons and multitons.
Finally in this part, build '''hcllevels''' and '''hcllayerwzero'''. For hcllevels we are going to compute mean distances between clusters. It is computational infeasible to do this for all layers. And then for all layers (inc zero) we are going to run our selection regression.
===Other===
