Sentiment plot: diverging bar, or a positive/negative comparison cloud

A ggplot2 diverging bar for a tidy_sentiment result by default: positive counts extend one direction, negative counts the other, so bar position (not colour alone) carries the primary polarity signal, the same convention sframe_draw_likert_diverging() uses for Likert agreement (dark ramp toward the pole) rebuilt here in ggplot2 rather than called directly, since that helper is base-graphics and Likert-scale-specific. Facets by group when result$table carries a group column, mirroring sframe_plot_term_frequency()'s grouped branch.

Usage

sframe_plot_sentiment(result, palette = c("web", "print"))

Arguments

result

A tidy_sentiment result list from run_analysis_plan().

palette

One of "web" or "print". See sframe_brand().

Value

A ggplot2 object, or NULL when the result carries no table.

Details

When result$options$wordcloud is TRUE (opt-in, default FALSE, matching sframe_plot_term_frequency()'s own word-cloud toggle), draws a comparison cloud instead: negative-sentiment words above the centre line, positive-sentiment words below it, each word sized by how often it occurred, using the internal tidy_sentiment runner's $word_sentiment word-by-sentiment counts. Answers a different question from the diverging bar: not "how many responses leaned positive," but "which words drove that."

See also

run_analysis_plan(), sframe_draw_likert_diverging()