library(topicmodels)
rv_dtm <- tidytext_woord %>%
group_by(bestandsnaam, text) %>%
summarise(aantal = n()) %>%
cast_dtm(term = text, document = bestandsnaam, value = aantal)`summarise()` has grouped output by 'bestandsnaam'. You can override using the
`.groups` argument.rv_lda <- LDA(rv_dtm, k = 10, control = list(seed = 1234))rv_topics <- tidy(rv_lda, matrix = "beta")
rv_topics# A tibble: 139,230 × 3
topic term beta
<int> <chr> <dbl>
1 1 1 0.00635
2 2 1 0.00268
3 3 1 0.00556
4 4 1 0.00397
5 5 1 0.00797
6 6 1 0.0112
7 7 1 0.00225
8 8 1 0.00729
9 9 1 0.00966
10 10 1 0.00430
# ℹ 139,220 more rowsrv_top_terms <- rv_topics %>%
group_by(topic) %>%
slice_max(beta, n = 10) %>%
ungroup() %>%
arrange(topic, -beta)
rv_top_terms %>%
mutate(term = reorder_within(term, beta, topic)) %>%
ggplot(aes(beta, term, fill = factor(topic))) +
geom_col(show.legend = FALSE) +
facet_wrap(~ topic, scales = "free") +
scale_y_reordered()