Visualize bigram chains
Usage
plot_bigrams(
df,
feature = word,
random_seed = TRUE,
set_seed = NULL,
legend = FALSE,
top_n = 35
)Arguments
- df
A tidy data frame potentially containing a column called "word" or columns called "word_1" and "word_2".
- feature
The feature to use when constructing ngrams
- random_seed
Whether to randomize the creation of the network chart.
- set_seed
A specific seed to use if not random
- legend
Whether to show a legend for the edge color
- top_n
The number of pairs to visualize
See also
Other visualizing helpers:
change_colors(),
plot_doc_word_bars(),
plot_doc_word_heatmap(),
plot_hapax(),
plot_htr(),
plot_tf_idf(),
plot_topic_bars(),
plot_topic_distributions(),
plot_topic_wordcloud(),
plot_ttr(),
plot_vocabulary(),
visualize()
Other n-gram helpers:
add_ngrams(),
combine_ngrams(),
separate_ngrams()
Examples
if (FALSE) { # \dontrun{
# It isn't necessary to use add_ngrams()
df |>
plot_bigrams()
# Adding them first allows for filtering steps
df |>
add_ngrams() |>
drop_stopwords(word_1) |>
drop_stopwords(word_2) |>
plot_bigrams()
# Only bigrams influence the visualization These show the same networks:
df |>
add_ngrams() |>
plot_bigrams()
df |>
add_ngrams(4) |>
plot_bigrams()
} # }
dubliners <- get_gutenberg_corpus(2814) |>
load_texts() |>
identify_by(part) |>
standardize_titles()
dubliners |>
plot_bigrams()
# Loading `ggraph` enables edge to show connection strengths
library(ggraph)
dubliners |>
plot_bigrams()
dubliners |>
add_ngrams(2) |>
drop_stopwords(feature = word_1) |>
drop_stopwords(feature = word_2) |>
plot_bigrams()
dubliners |>
dplyr::filter(doc_id == "The Dead") |>
plot_bigrams(top_n = 70) |>
change_colors(c("black", "orange"))
#> Scale for edge_colour is already present.
#> Adding another scale for edge_colour, which will replace the existing scale.
