I have a function in R recipe that creates bag of words dataframe. I would like to put all the substrings (bagofwords$text) into MULTISELECT choices. However, since I feel way more comfortable with R than Python I can't think of the solution.
Thank you for your answer. Well maybe there is other way to solve this problem? to be precise I would like to use these substring as parameter for str_replace_all. I want to drop most frequant substrings from initial string column, because it would improve my trained model prediction accuracy and would reduce noise.