termco
termco copied to clipboard
trinker
Regular Expression Counts of Terms and Substrings
``` ## convert names columns with one hot encoding to list of column names tags2list % select(...) %>% {Map(function(x, y){ ifelse(x > 0, y, NA) }, ., names(.))} %>% unlist(use.names...
add heirachical elements to a list. Also add a check for a heirchy with a vector of vector regexes which causes a warninging (should be an error)
If **quanteda** provides a means to convert to dfm from dtm use that for easier maintainability: https://github.com/quanteda/quanteda/issues/1222
As I work through the process of expert rules regex tagging I want to get out some theory and terms so that they can be iterated on and used in...
https://gist.github.com/trinker/5b3b1932d59594bb131958c6dd6fdf04 Would want to use textplot instead or pull the guts out into a separate package that feeds them both minimally.
This is out of date for current thinking and new methods
Take 2 validate models and overwrite the first with info from the second so long as the sample taken was the same.