Time elapsed between consecutive first-time authors, by source and popularity stratum

Event-level alternative to new_author_rate_tbl()'s repo-month rate: instead of asking "how many new arrivals landed this month, normalised by repo-months of exposure", this asks "how long does a repository wait between one first-time author and the next". For one source, orders each repository's distinct issue authors by their own first-ever appearance in that repo's issue tracker (its founding author first, as identified in new_author_rate_tbl() - but kept here rather than excluded, since it anchors the very first interval) and computes, for every author after the founder, the elapsed time since the previous first-time author's own first appearance. Each interval is time-stamped at the arriving author's own issue, not at the interval's start or midpoint, so a repository that goes quiet for two years and then gains a new contributor logs one long interval dated to the day that contributor actually showed up, not smeared backward across the quiet period. Needs no repo-months exposure denominator at all - a repository with only one author so far simply contributes no interval, rather than a zero.

Usage

author_interval_tbl(
  issue_authors_tbl,
  repo_tbl,
  source_name,
  n_strata = 4L,
  date_start = as.Date("2015-01-01"),
  date_end = NULL
)

Arguments

issue_authors_tbl

As returned by fetch_issue_authors().

repo_tbl

As returned by build_repo_tbl().

source_name

One of repo_tbl$source ("pypi", "npm", "joss", "ropensci").

n_strata

Number of popularity strata.

date_start, date_end

Date bounds on which authors' first appearances are considered; unlike the repo-month rate tables, these bound the raw event timestamps directly rather than a floored calendar month, and date_end defaults to Sys.Date() (today), not the start of the current month, since there is no partial-month repo-months exposure to worry about truncating here.

Value

A tibble with one row per (repository, author arrival after the founder): repo_url, popularity_stratum, arrival_index (2 = the first author after the founder, 3 = the second, and so on), event_time (the arriving author's own first-issue timestamp - when this interval is "logged"), interval_days (elapsed time, in days, since the previous first-time author's own first appearance). Carries source_name as an attribute.

Examples

if (FALSE) { # \dontrun{
iv <- author_interval_tbl (issue_authors_tbl, repo_tbl, "pypi")
plot_author_interval (issue_authors_tbl, repo_tbl)
} # }