Term frequency for open-ended text

Tokenises text (splitting on whitespace, lower-casing, and stripping punctuation), removes stop words, and counts term frequency. Ships a small built-in English stop-word list so this works with zero optional packages; pass stop_words = character(0) to disable filtering, or a custom vector to override it.

Usage

term_frequency(text, stop_words = NULL, top_n = 30L)

Arguments

text

Character vector of responses (raw or already cleaned by clean_text_responses()).

stop_words

Character vector of words to exclude, or NULL to use the built-in English list, or character(0) for no filtering.

top_n

Integer. Maximum number of terms to return, most frequent first. Default 30.

Value

A data.frame with columns term, n, and pct.

Examples

demo <- sframe_demo_data()
cleaned <- clean_text_responses(demo$responses, "comments")
head(term_frequency(cleaned, top_n = 10))
#>             term  n  pct
#> 1        details 35 13.0
#> 2           like 35 13.0
#> 3 sustainability 35 13.0
#> 4          clear 31 11.5
#> 5        content 31 11.5
#> 6         online 31 11.5