Compare monthly issue rate across all four sources (pypi, npm, joss, ropensci), for one popularity stratum. Note that "stratum" is relative to each source's own distribution (see issue_rate_tbl()/ popularity_strata()) - e.g. pypi's Q4 download count and joss's Q4 star count aren't the same absolute popularity, just each source's own top quarter. A source with no data yet for the requested window (e.g. not fully fetched - see analysis-plan.md) just contributes no line, rather than erroring.

Compare monthly issue rate across all four sources (pypi, npm, joss, ropensci), for one popularity stratum. Note that "stratum" is relative to each source's own distribution (see issue_rate_tbl()/ popularity_strata()) - e.g. pypi's Q4 download count and joss's Q4 star count aren't the same absolute popularity, just each source's own top quarter. A source with no data yet for the requested window (e.g. not fully fetched - see analysis-plan.md) just contributes no line, rather than erroring.

Usage

plot_activity_by_source(
  issue_authors_tbl,
  repo_tbl,
  stratum,
  n_strata = 4L,
  contrib_threshold = 0.01,
  metric = c("issues", "comments"),
  window = 12L,
  relative = TRUE,
  start_year = NULL,
  ros_joss_mult = 20
)

Arguments

issue_authors_tbl

As returned by fetch_issue_authors().

repo_tbl

As returned by build_repo_tbl().

stratum

Integer popularity stratum to compare (1 = lowest popularity, n_strata = highest), matching one of issue_rate_tbl()'s popularity_stratum levels ("Q<stratum>").

n_strata, contrib_threshold, metric, window

Passed to each source's issue_rate_tbl() call; n_strata must be the same one stratum is a level of.

relative

If TRUE (default), rescale each source by its own mean before plotting - sources sit on very different absolute rate scales (e.g. pypi's raw issue traffic dwarfs ropensci's), which would otherwise squash the smaller sources' trends to flat lines near zero. Puts every line at a comparable "around 1 = that source's own average" scale, so trends are comparable even though absolute rates aren't. Set FALSE to plot absolute rates instead.

start_year

Optional year (e.g. 2018) to start the analysis from - passed straight through as each issue_rate_tbl() call's date_start, so it also governs the repo-months/rate calculations themselves, not just the plotted range. NULL (default) starts from 2015-01-01. There is no equivalent end-date control - analyses always run up to the current month.

ros_joss_mult

Multiplier applied to the final (post-relative) rate values for the "ropensci" and "joss" sources only, after every other calculation. Default 10. This is a display-only scaling of those two sources relative to "pypi"/"npm" - the plot's y-axis label is annotated whenever it's not 1, so the scaling isn't silently hidden from anyone reading the plot.

Value

A ggplot object.

Examples

if (FALSE) { # \dontrun{
plot_activity_by_source (issue_authors_tbl, repo_tbl, stratum = 4L)
} # }