1
0
mw-lifecycle-analysis/commit_analysis/framework_commit_collation.R

90 lines
3.2 KiB
R

library(tidyverse)
library(purrr)
library(readr)
library(stringr)
library(lubridate)
library(tidyr)
data_dir = "/gscratch/comdata/users/mjilg/mw-repo-lifecycles/commit_data/bot_frameworks"
csv_files <- list.files(data_dir, pattern = "*.csv", full.names = TRUE)
read_and_label <- function(file) {
project_name <- basename(file) %>%
stringr::str_remove("_commits.csv")
read_csv(file) %>%
mutate(project = project_name)
}
all_data <- csv_files %>%
map_df(read_and_label)
# TODO: this is project/event specific
event_date <- as.Date("2013-07-01")
#event_date <- as.Date("2013-04-25")
#event_date <- as.Date("2012-12-11")
df <- all_data |>
mutate(commit_date = ymd_hms(commit_date))
df <- df %>%
group_by(project) %>%
mutate(oldest_commit_date = min(as.Date(commit_date))) %>%
ungroup() %>%
mutate(age = as.numeric(as.Date("2025-02-10") - oldest_commit_date))
filtered_df <- df %>%
group_by(project) %>%
filter(min(as.Date(commit_date)) <= event_date) %>%
ungroup()
calculated_start_date <- event_date %m-% months(12)
start_date <- max(calculated_start_date, df$oldest_commit_date)
end_date <- event_date %m+% months(12)
#getting the relative weeks to the publication date
relative_week <- function(date, ref_date) {
as.integer(as.numeric(difftime(date, ref_date, units = "days")) %/% 7)
}
filtered_df <- filtered_df |>
mutate(relative_week = relative_week(commit_date, event_date)) |>
arrange(relative_week) |>
group_by(author_email) |>
mutate(new_author = ifelse(row_number() <= 5, 1, 0),
new_author_wmf = if_else(grepl("@wikimedia", author_email), new_author, 0),
new_author_unaff = if_else(!grepl("@wikimedia", author_email), new_author, 0)) |>
ungroup()
weekly_commits <- filtered_df |>
group_by(project, relative_week, age) |>
summarise(commit_count = n(),
author_emails = list(unique(author_email)),
committer_emails = list(unique(committer_email)),
mediawiki_dev_commit_count = sum(grepl("@users.mediawiki.org", author_email)),
wikimedia_commit_count = sum(grepl("@wikimedia", author_email)),
wikia_commit_count = sum(grepl("@wikia-inc.com", author_email)),
bot_commit_count = sum(grepl("l10n-bot@translatewiki.net|tools.libraryupgrader@tools.wmflabs.org", author_email)),
wmf_ft_commit_count = sum(new_author_wmf),
unaff_ft_commit_count = sum(new_author_unaff),
.groups = 'drop') |>
replace_na(list(commit_count = 0)) |>
replace_na(list(wikimedia_commit_count = 0)) |>
replace_na(list(l10n_commit_count = 0)) |>
replace_na(list(jenkins_commit_count = 0)) |>
replace_na(list(mediawiki_dev_commit_count = 0)) |>
replace_na(list(wikia_commit_count = 0)) |>
replace_na(list(wmf_ft_commit_count = 0)) |>
replace_na(list(unaff_ft_commit_count = 0)) |>
mutate(before_after = if_else(relative_week < 0, 0, 1)) |>
select(-author_emails, -committer_emails)
weekly_commits <- weekly_commits |>
filter(relative_week >= (-52) & relative_week <= 52 )
weekly_commits
output_filepath <-"/mmfs1/gscratch/comdata/users/mjilg/mw-repo-lifecycles/case1/event_0314_bot_frameworks_weekly_commit_count_data.csv"
write.csv(weekly_commits, output_filepath, row.names = FALSE)