Plot step-change in non-core rate across sources and strata

Bar chart of step_change_tbl() output: one column per source, one bar per popularity stratum, step_change on a log y-axis (so a halving and a doubling are visually symmetric) with a reference line at 1 (no change). If tbl carries a metric column (e.g. issues vs. comments, row-bound from two step_change_tbl() calls), it is filtered down to the single metric requested before plotting - one call produces one single-metric plot, so issues and comments are two separate figures rather than two facet rows of the same one.

Usage

plot_step_change(
  tbl,
  source_display = NULL,
  metric = c("issues", "comments"),
  ref_date = as.Date("2021-01-01")
)

Arguments

tbl: As returned by step_change_tbl(), optionally with an added metric column.
source_display: Named character vector mapping internal source names to display labels, e.g. SOURCE_DISPLAY_NAME.
metric: Which metric to plot: "issues" (default) or "comments". Only used to filter tbl down to one metric when it carries a metric column (matched case-insensitively against that column's values, e.g. "Issues"/"Comments"); otherwise tbl is assumed to already be single-metric, and this only sets the plot title.
ref_date: The same ref_date passed to step_change_tbl(), used only to label the plot.

Value

A ggplot object.

Examples

if (FALSE) { # \dontrun{
fc <- step_change_tbl (issue_authors_tbl, repo_tbl, c ("cran", "npm"))
plot_step_change (fc, SOURCE_DISPLAY_NAME, metric = "issues")
} # }