Open stephbuon opened 3 years ago
@EliasLMann -- here is the other place we would use the LL implementation that you wrap.
@EliasLMann :
steps would be something like:
1) check out this existing C++ implementation: http://cahillsoftware.com/media/af21843f135214fffff850fffffe417.html
Is it any good (i.e. will it return results in real time)? If not, it is reasonable to improve it for our app? (remember, we want to measure perhaps two speakers for a single decade.)
2) figure out how to call the C++ implementation in R using rcpp: https://cran.r-project.org/web/packages/Rcpp/index.html
3) figure out how to add this implementation to our application
Add Log Likelihood to the Sentiment Laden Collocates Panel