better handling of cluster internals

oturns / geosnap

The Geospatial Neighborhood Analysis Package

https://oturns.github.io/geosnap-guide

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

247 stars 32 forks source link

better handling of cluster internals #123

Closed knaaptime closed 5 years ago

knaaptime commented 5 years ago

this PR makes a few improvements to the clustering internals

it uses multiindices rather than creating several hacky columns then dropping them
it adds a new scaler argument that accepts any sklearn.preprocessing class to scale inputs
more flexible handling of spatial weights, including allowing users to provide their own (i.e. allow distance-based weights for algos that can support them)

knaaptime commented 5 years ago

these tests are passing now

renanxcortes commented 5 years ago

I think in the build is raising a FileNotFoundError: [Errno 2] No such file or directory: '/home/travis/build/spatialucr/geosnap/ltdb_sample.zip' is this expected?

knaaptime commented 5 years ago

i'm not sure why, but it's because the branch doesnt have the environment variables set or something. You can see that it passes on my fork

knaaptime commented 5 years ago

(if it's merged, the tests will pass on spatialucr/master)

renanxcortes commented 5 years ago

i'm not sure why, but it's because the branch doesnt have the environment variables set or something. You can see that it passes on my fork

yep, I see that.... weird

renanxcortes commented 5 years ago

(if it's merged, the tests will pass on spatialucr/master)

Ok, I took a quick look at the code but, most importantly, read the bullet points of the PR and it is fine to me.