search

cjbayesian / ml4h_paper_2019

Data Pipeline for analysis of ML4H 2018 workshop at NeurIPS.
4 stars 0 forks source link

Build example LDAvis #4

Closed cjbayesian closed 5 years ago

cjbayesian commented 5 years ago

Create an example topic model and visualization to serve as a first pass.

cjbayesian commented 5 years ago

See https://github.com/cjbayesian/ml4h_paper_2019/commit/f3194a423591cdb2ff1fc8c66f87c8ebb0433d64

which generates a simple topic model of the ML4H papers and creates the interactive visualization: image

beamandrew commented 5 years ago

Nice! How did you select the number of topics?

Could you throw it up on shinyapps.io if it's not too much trouble?

cjbayesian commented 5 years ago

I haven't used shinyapps.io, but I put the files on the google drive here https://drive.google.com/drive/folders/1pb5PGPLNNquWZXnSqLh24WEzzbN436Ne

You can spin it up locally really easily by downloading that folder and running 

python -m SimpleHTTPServer 8000

from that folder then opening localhost:8000 in your browser.

I just used all the default parameters from the example in the LDAvis docs.