release prep - Githubissues

pysal / segregation

Segregation Measurement, Inferential Statistics, and Decomposition Analysis

https://pysal.org/segregation/

BSD 3-Clause "New" or "Revised" License

112 stars 26 forks source link

release prep #84

Closed knaaptime closed 5 years ago

knaaptime commented 5 years ago

still working a bit onthe readme, but checking this in as it resolves the last of the release checklist

renanxcortes commented 5 years ago

This is absolutely awesome so far! So, specifically for the ReadMe, I think we need to address two things:

Put the Decomposition title before the paragraph that starts "Another useful analytics that can be performed..."
Rename the second notebook link in the readme from using euclidian distances to something more generic. Only a specific set of indexes uses euclidian distances and, also, the nature of some indexes do not even use the concept of distance. So, one option for this link would be using vector-based spatial data.

Regarding notebooks, sphinx documentation and tests for the new names SingleValueTest and TwoValueTest I can do this after merging this PR, ok?

knaaptime commented 5 years ago

+1, there's another big commit coming on the readme

renanxcortes commented 5 years ago

Also, I was thinking here that we can also address that DOI badge in this PR :)

knaaptime commented 5 years ago

Unfortunately that will have to be the next one. The badge doesn't get created until the release. Will do a small PR after the bump that only updates the readme

renanxcortes commented 5 years ago

Unfortunately that will have to be the next one. The badge doesn't get created until the release. Will do a small PR after the bump that only updates the readme

Got it!

renanxcortes commented 5 years ago

Ok, cool, this ReadMe got really pleasing.

But I still think some 3 things need to be addressed:

Actually, using spatial weights matrices & euclidian distances don't reflect the nature of these set of indexes. For example, The Boundary Spatial Dissimilarity uses the shared boundary of two sets of polygons, the perimeter area index uses the ratio of the perimeter and area, all concentration indexes (DEL, ACO, RCO) deals with the polygon area and sorting them according to the area. All of those indexes (and other we might add) do not use a spatial weight matrices.
We could also use the income example call for SpatialDissim to be harmonized with the previous call:

spatial_index = SpatialDissim(gdf, "hi_income", "total_population")

Actually, we do not require GeoDataFrame for the MultiGroup Measures so the sentence though note, again, that now we requre a GeoDataFrame for proper estimation can be dropped.

knaaptime commented 5 years ago

Actually, using spatial weights matrices & euclidian distances don't reflect the nature of these set of indexes. For example, The Boundary Spatial Dissimilarity uses the shared boundary of two sets of polygons, the perimeter area index uses the ratio of the perimeter and area, all concentration indexes (DEL, ACO, RCO) deals with the polygon area and sorting them according to the area. All of those indexes (and other we might add) do not use a spatial weight matrices.

that shared boundary is calculated with... a spatial weights matrix :P

renanxcortes commented 5 years ago

that shared boundary is calculated with... a spatial weights matrix :P

Hehehehe, you're right, I didn't pay attention to that! Everything ok with the last commit! I think this is good to go! Thank you!!