R/df2token.R

In jaytimm/text2df:

#' Convert a df column to token object by group(s).
#'
#' @name df2token
#' @param df a dataframe
#' @param token Colname to be tokenized, as string
#' @param by Colname(s) to group token object, as string
#' @return A list (or tokens object)
#'
#' @export
#' @rdname df2token
#'
#'
df2token <- function(df,
                     token,
                     by = c('doc_id')){

  nn <- lapply(1:length(by), function(x) df[[by[x]]])
  df$id99 <- do.call("paste", c(nn, sep=":"))
  split(df[[token]], df$id99)

  }