doug-friedman / topicdoc Goto Github PK
View Code? Open in Web Editor NEWTopic-Specific Diagnostics for LDA and CTM Topic Models
Home Page: https://doug-friedman.github.io/topicdoc/
License: Other
Topic-Specific Diagnostics for LDA and CTM Topic Models
Home Page: https://doug-friedman.github.io/topicdoc/
License: Other
This is more of a feature request rather than an issue.
I was wondering if you are planning to implement any support for model diagnostics of topic models in embedded spaces. The paper I am referring to can be found here. There is an implementation of the original code in R here.
It seems that the R package does not support any of such things that seem to be well implemented in the original Python code.
Thanks!
Prepare for release:
usethis::use_cran_comments()
devtools::check()
devtools::check_win_devel()
rhub::check_for_cran()
Submit to CRAN:
devtools::submit_cran()
Wait for CRAN...
usethis::use_github_release()
usethis::use_dev_version()
usethis::use_news_md()
In the function coherence
, shouldn't it be for (l in 1:(m - 1))
instead of for (l in 1:(top_n_tokens - 1))
? Otherwise, you are counting some but not all of the pairs of words twice. This is also the definition in Mimno et. al. 2011, p. 265.
The current workflow is a bit out of date so migrating this github actions should make things smoother.
My initial thoughts are that this package should support the existing diagnostics for structural topic models fit using the stm library. Maybe give the user the option to use that library's diagnostic calculations (e.g. semanticCoherence
) if they'd like?
Right now the github page for this package deploys via other CI/CD services and should deploy via github actions.
The R-CMD-check Github action already covers this so travis-ci and circleci are just duplicative.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.