Exploring external contributions to the R codebase used by Roche to design and analyse late-stage clinical trials
R is increasingly used in the pharmaceutical industry as the backbone for the pan-study codebase for the design and analysis of clinical trials. In parrallel with this shift to R, many companies are open sourcing, and collaborating, on the post-competitive code used across studies. The Pharmaverse and openstatsware are two example initiatives for statistical programming, and biostatistics, respectivly.
Whlie numerous benefits come from companies open sourcing their R codebase, from better talent acquisition, to transperancy with regulators, activity on git repos provides an insight into the return on investment (ROI) from external contributions to the codebase a company depends on. In this document we explore the ROI as assessed via external contributions to the late-stage codebase at Roche, shedding light on the tangible benefits derived from collaborative development in the pharmaceutical domain.
Open Source, Pharmaverse, openstatsware
This is a working document, and is not yet ready for public consumption.
In [1]:
library(glue)
library(dplyr)
Attaching package: 'dplyr'The following objects are masked from 'package:stats':
filter, lagThe following objects are masked from 'package:base':
intersect, setdiff, setequal, unionlibrary(GitStats)
library(gt)
library(gtExtras)
library(ggplot2)
library(git2r)
Attaching package: 'git2r'The following object is masked from 'package:GitStats':
resetThe following object is masked from 'package:dplyr':
pulllibrary(tidyr)In [2]:
new_data <- FALSE
# Google sheets data ----------------------------------------------------------
# gsheet id
sheet_id <- "1EV145GdYKTuYkOvwxIeWIINwGZkb-oi6ZLCq7d5xFRM"
# Roche people, alongside their github handles
roche_people <- googlesheets4::read_sheet(sheet_id, "roche-peeps")ℹ Suitable tokens found in the cache, associated with these emails:• 'james.black.jb2@roche.com'• 'james@epijim.uk' Defaulting to the first email.! Using an auto-discovered, cached token. To suppress this message, modify your code or options to clearly consent to
the use of a cached token. See gargle's "Non-interactive auth" vignette for more details: <https://gargle.r-lib.org/articles/non-interactive-auth.html>ℹ The googlesheets4 package is using a cached token for
'james.black.jb2@roche.com'.✔ Reading from "Roche Pharmarverse contributors + packages".✔ Range ''roche-peeps''. roche_pharmaverse <- googlesheets4::read_sheet(
sheet_id, "roche-pharmaverse"
) |>
mutate(
full_name = glue("{org}/{name}")) |>
mutate(
url_repo = glue("https://github.com/{full_name}"),
url_hex = case_when(
!is.na(logo_path) ~ glue(
"https://github.com/{full_name}/{logo_path}"
)
)
) ✔ Reading from "Roche Pharmarverse contributors + packages".✔ Range ''roche-pharmaverse''. # roche github handles as vector
roche_gh_handles <- na.omit(roche_people$gh_username)
# Roche people, via the emails in their gitconfig
roche_gitconfig <- googlesheets4::read_sheet(sheet_id, "git-author") |>
na.omit() |>
left_join(
roche_people,
by = c("gh_username")
) ✔ Reading from "Roche Pharmarverse contributors + packages".✔ Range ''git-author''.New names:
• `` -> `...3`
• `` -> `...4`if (new_data){
## Git Stats -----------------------------------------------------
# git_stats <- create_gitstats() %>%
# set_host(
# api_url = "https://api.github.com",
# orgs = {
# roche_pharmaverse |>
# filter(platform == "github.com") |>
# pull(org) |>
# unique()
# }
# ) %>%
# set_host(
# api_url = "https://code.roche.com.com/api/v4",
# orgs = {
# roche_pharmaverse |>
# filter(platform == "code.roche.com") |>
# pull(org) |>
# unique()
# }
# )
#
# ## All repos
# pull_repos(git_stats)
# pull_commits(git_stats, date_from = as.Date("2010-01-01"), date_until = Sys.Date() )
#
# repos <- get_repos(git_stats)
# commits <- get_commits(git_stats)
## Git Commit Stats -----------------------------------------------------------
all_commits <- GithubMetrics::gh_commits_get(
paste0(roche_pharmaverse$org,"/",roche_pharmaverse$name),
days_back = 365*10
)
cleaned_commits <- all_commits |>
# Add windows
left_join(
roche_people |>
filter(!is.na(gh_username)),
by = c("author" = "gh_username")
) |>
# Just packages from Roche
filter(full_name %in% roche_pharmaverse$full_name) |>
# Is a Roche person
mutate(
commit_date = as.Date(datetime),
# if only start date, set last day to tomorrow
`Roche end` = case_when(
!is.na(`Roche start`) & is.na(`Roche end`) ~ Sys.Date() + 1,
TRUE ~ `Roche end`
),
# if only end date, set start day to earliest commit
`Roche start` = case_when(
is.na(`Roche start`) & !is.na(`Roche end`) ~
min(as.Date(all_commits$datetime)),
TRUE ~ `Roche start`
),
# No set if commit is Roche or not
is_ours = case_when(
is.na(`Roche start`) &
tolower(author) %in% tolower(roche_gh_handles) ~ "Roche",
commit_date > `Roche start` & commit_date < `Roche end` &
tolower(author) %in% tolower(roche_gh_handles) ~ "Roche",
grepl("roche.com", commit_email, fixed = TRUE) ~ "Roche",
TRUE ~ "External"
)
) %>%
# Add which namespace
left_join(
roche_pharmaverse |>
select(full_name, namespace),
by = "full_name")
saveRDS(cleaned_commits, "cache/cleaned_commits.rds")
gh_usernames_status <- cleaned_commits |>
select(author, is_ours) %>%
unique()
## Who is there
# cleaned_commits |>
# group_by(author, full_name, is_ours) |>
# summarise(n = n()) |> unique()
# Git lines of code -----------------------------------------------------------
# Pull repos down and do a git blame
repo <- roche_pharmaverse$full_name[1]
data_lines <- NULL
for(repo in roche_pharmaverse$full_name){
message(repo)
path <- file.path(tempfile(pattern=repo))
dir.create(path, recursive=TRUE)
#output <- clone(glue("https://github.com/{repo}"), path, progress = FALSE)
system(glue("git clone https://github.com/{repo} {path}"))
system(glue("cd {path} && {getwd()}/contributors.sh"))
file.copy(from = glue("{path}/temp.csv"), to = "temp.csv",overwrite = TRUE)
temp <- readr::read_table("temp.csv", col_names = FALSE)
i_data <- bind_cols(
temp %>%
select(lines = X1),
temp %>%
select(-X1,-X2) |>
tidyr::unite("person",1:(ncol(temp)-2),sep=" ") |>
mutate(
person = gsub(x = person,replacement = "",pattern = " NA",ignore.case = FALSE)
)
) |>
mutate(repo = repo)
data_lines <- bind_rows(
i_data, data_lines
)
}
saveRDS(data_lines,"cache/cleaned_lines.rds")
## Get issues data from github ------------------------------------------------
issues <- GithubMetrics::gh_issues_get(
roche_pharmaverse$full_name,
days_back = 30*12
)
comments <- GithubMetrics::gh_issues_comments_get(
roche_pharmaverse$full_name,
days_back = 30*12)
issues_cleaned <- left_join(
comments %>%
select(
full_name,
issue_number,
comment_created = created,
comment_body = body,
comment_author = author
),
issues,
by = c("full_name","issue_number")
) |>
filter(comment_author != "github-actions[bot]") |>
mutate(
Package = basename(full_name)
) |>
left_join(
roche_pharmaverse |> select(name, namespace),
by = c("Package" = "name")
)
} else {
data_lines <- readRDS("cache/cleaned_lines.rds")
cleaned_commits <- readRDS("cache/cleaned_commits.rds")
issues_cleaned <- readRDS("cache/issues_cleaned.rds")
}In July 2021, Roche stopped development with propriety statistical software, to focus on a new backbone of R packages for the analysis of clinical trials. A 10+-year old codebase written in a propriety language (named STREAM) went into maintenance only updates, and development resources were shifted in their entirety to the new R backboned codebase, that comprised rOAK, admiral (Straub et al. 2023) and NEST (NEST 2023), which form the core of the pharmaverse(pharmaverse 2023). The design of clinical trials and exploratory data analysis at Roche has a longer history of R use, with a packages like rpact (Anders Bilgrau and Krøgholt 2023) and crmPack (Sabanés Bové et al. 2019) used for many years. This has continued to increase in recent years through initiatives like openstatsware (openstatsware 2023), that aim to collaboratively fill software gaps in clinical trial design as open source software.
In this document we explore the ROI from the perspective of an organisation, from both qualitive and quantative assessments, shedding light on the tangible benefits derived from collaborative development in the pharmaceutical domain.
Regardless of whether a company open sources it’s own code, with our industries away from proprietary languages we are likely to be both depending on and extending open source software. A core question is then whether there is an added benefit open sourcing our own code, and actively contributing back to projects we use.
The PHUSE Open Source Guidance (PhUSE 2023) document outlines the benefits of open source across 3 main points:
Post-competitive intellectual property is a term defined in the guidance as:
A less common term we have defined to be where code collaboration improves the efficency of insights, rather than the creation of insights that would otherwise not be possible. In the context of PHUSE collaborators, this includes packages that take CDISC data and apply templated data steps and visualizations to prepare a CSR, like those seen in the pharmaverse.
A rising tide lifts all boats embodies a position Roche took in the early days of open sourcing, with the specific intent that by open sourcing our code, and collaborating with companies, we could help regulators and our industry gain confidence in the use of open source software regardless of whether we saw external contributions back onto the code we released.
This was a strategy that worked across three layers;
pharmaverse packages like admiral and NEST, and active investment co-founding R/Pharma and it’s non-profit Open Source in Pharma (Pharma 2023)R Consortium (Consortium 2023), and the first Pharma to co-host useR!, the main global R conference.The PHUSE Open Source Guidance (PhUSE 2023) document highlights several risks from open sourcing code. Many of these can be directly countered, as shown in Table 1.
| Risk statement | Rebuttal |
|---|---|
| We could be liable to end users | All licences Roche use have provisions in them that explicity state the code is |
| Opening up our code could risk our reputation if bugs are found | Ultimately this would be in the benefit of patients if we are made aware of ways to improve the code |
| Open sourcing is an unnecessary risk |
|
Another in-development complexity is the Cyber Resilience Act (Commission 2022), currently being discussed in the European Union. While we currently only have a draft of the new law, it is likely to have implications for the use and contributions to open source software by pharma companies.
Examples include;
In [3]:
n_pkgs_pharmaverse <- roche_pharmaverse |>
filter(namespace == "pharmaverse") |>
nrow()
n_pkgs_openstatsware <- roche_pharmaverse |>
filter(namespace == "openstatsware") |>
nrow()
n_roche_people <- n_distinct(roche_people$name)
n_roche_accounts <- nrow(roche_people)As the code-base is version controlled with git, we are able to extract some information around who has made contributions, and what types of contributions those were. There are 20 packages in the pharmaverse relevant to Roche, and 6 packages from openstatsware. Across these packages, we have 203 Roche people working across 205 known Github user accounts.
In this analysis we can look at commits ( Section 3.2), lines of code (Section 3.3) and issues/commits (Section 3.4) to understand better how our code-base is influenced by individuals from outside of our own company.
We maintain a list of Github handles of Roche employees used to contribute to pharmaverse packages we consider relevant to submissions within Roche or openstatsware packages used for statistical design or analysis.
The following known caveats exist with this analysis.
devtools::document(), and so certain file-types are ignored) is automatically adding the majority of the code in the repo.Figure 1 shows.
In [4]:
cleaned_commits |> ungroup() |>
mutate(
Month = as.Date(paste0(format(as.Date(datetime), "%Y-%m"),"-01"))
) |>
group_by(namespace, Month) %>%
summarise(
commits_roche = sum(is_ours == "Roche"),
commits_all = n(),
.groups = "drop"
) |> ungroup() |>
filter(commits_all >= 10) |>
mutate(
`% commits from Roche` = commits_roche/commits_all
) |>
ggplot(aes(x=Month, y=`% commits from Roche`)) +
geom_line(alpha = 0.3)+
geom_point(aes(size = commits_all), alpha = 0.5) +
ggthemes::theme_hc() +
facet_grid(rows = vars(namespace)) +
scale_y_continuous(labels = scales::percent_format(accuracy = 1)) +
labs(
title = "Monthly % of commits from Roche",
subtitle = "Months with <10 total commits excluded",
size = "Total commits"
)
Figure 1:
cleaned_commits |>
group_by(namespace, full_name) |>
summarise(
`Roche commits` = sum(is_ours == "Roche"),
`External commits` = sum(is_ours == "External"),
Commits = n(),
`First commit` = min(datetime),
Contributors = n_distinct(author),
`Roche contributors` = n_distinct(author[is_ours == "Roche"]),
.groups = "drop_last"
) %>%
mutate(
Package = basename(full_name),
`% Commits` = `Roche commits`/Commits,
`% Contributors` = `Roche contributors`/Contributors,
`Age (y)` = round((Sys.Date() - as.Date(`First commit`))/365.25,1)
) |>
left_join(
roche_pharmaverse |>
select(name, url_repo, url_hex),
by = c("Package" = "name")
) |>
left_join(
cleaned_commits |>
mutate(Monthly = format(as.Date(datetime), "%Y-%m") ) |>
group_by(full_name, Monthly) |>
summarise(
`Roche commits` = sum(is_ours == "Roche"),
Commits = n(),
.groups = "drop_last"
) %>%
mutate(
Package = basename(full_name),
`% Commits` = `Roche commits`/Commits
) |>
arrange(Monthly) %>%
summarise(
`Monthly trend` = list(`% Commits`), .groups = "drop"
) |>
ungroup() |>
mutate(
Package = basename(full_name)
) |>
select(
Package, `Monthly trend`
),
by = "Package"
) |>
mutate(
Package = dplyr::case_when(
!is.na(url_hex) ~ glue::glue(
"<img src='{url_hex}' height='24'></img> <a href='{url_repo}'>`{Package}`</a>"),
TRUE ~ glue("<a href='{url_repo}'>`{Package}`</a>")
)
) |>
select(
Source = namespace,
Package,
`Age (y)`,
Contributors,
Commits,
`% Contributors`,
`% Commits`,
`Monthly trend`
) |>
arrange(
desc(`% Commits`)
) |>
gt() |>
data_color(
columns = c(`% Contributors`,`% Commits`),
#rows = currency < 50,
method = "numeric",
palette = c("#F2D4FF","#007AC2"),
domain = c(0, 1)
) |>
tab_spanner(
label = "Roche metrics",
columns = c(`% Contributors`,`% Commits`, `Monthly trend`)
) |>
tab_spanner(
label = "Roche + Externals",
columns = c(Contributors, Commits)
) |>
tab_header(
title = "Contributions to the Roche relevant pharmaverse",
subtitle = glue("Only includes commit data for now, other forms of contribution will be added (e.g. issues)")
) |>
tab_source_note(source_note = md(
glue("Our list of github handles of Roche employees may not be complete. Our data on when employed (if they joined or left Roche) is manually added, so may not be exact. Generated with {nrow(cleaned_commits)} commits from {as.Date(min(cleaned_commits$datetime))} to {as.Date(max(cleaned_commits$datetime))}.")
)) |>
tab_options(
row_group.background.color = "#FAC9B5"
) |>
gt_plt_sparkline(
`Monthly trend`,
label = FALSE, same_limit = TRUE,
palette = c("black", rep("transparent", 4))
) |>
fmt_percent(
columns = c(`% Contributors`,`% Commits`),
decimals = 0
) |>
tab_footnote(
footnote = md("Person counted if ever employed by Roche, so contributions could occur while employed externally."),
locations = cells_column_labels(columns = `% Contributors`)
) |>
tab_footnote(
footnote = md("Where a person joined or left Roche, commits are assigned
to Roche only if employed when commit was made."),
locations = cells_column_labels(columns = `% Commits`)
) |>
tab_footnote(
footnote = md("Trend in % Roche comments month to month. X axis is not fixed across packages."),
locations = cells_column_labels(columns = `Monthly trend`)
) |>
fmt_markdown(columns = c(Package))| Contributions to the Roche relevant pharmaverse | ||||||
| Only includes commit data for now, other forms of contribution will be added (e.g. issues) | ||||||
| Package | Age (y) | Roche + Externals | Roche metrics | |||
|---|---|---|---|---|---|---|
| Contributors | Commits | % Contributors1 | % Commits2 | Monthly trend3 | ||
| pharmaverse | ||||||
| 6.7 | 4 | 252 | 100% | 100% | ||
| 6.3 | 26 | 1017 | 92% | 100% | ||
| 1.7 | 9 | 373 | 89% | 99% | ||
| 7 | 55 | 2809 | 93% | 95% | ||
| 5.4 | 30 | 1255 | 93% | 91% | ||
| 2 | 20 | 163 | 90% | 91% | ||
| 7 | 38 | 2666 | 89% | 91% | ||
| 6.5 | 48 | 1801 | 94% | 90% | ||
| 2 | 22 | 303 | 91% | 89% | ||
| 1.7 | 21 | 564 | 62% | 88% | ||
| 2 | 23 | 228 | 91% | 85% | ||
| 1.3 | 15 | 142 | 40% | 79% | ||
| 2.1 | 23 | 1000 | 48% | 78% | ||
| 3.1 | 74 | 4335 | 50% | 63% | ||
| 1.7 | 22 | 626 | 59% | 50% | ||
| 3.2 | 9 | 766 | 44% | 4% | ||
| 1.5 | 18 | 1505 | 22% | 3% | ||
| 3.1 | 8 | 369 | 12% | 1% | ||
| 3 | 4 | 87 | 0% | 0% | ||
| 2.2 | 3 | 160 | 0% | 0% | ||
| openstatsware | ||||||
| 2.8 | 8 | 1008 | 75% | 100% | ||
| 1.8 | 9 | 302 | 89% | 93% | ||
| 9.7 | 20 | 870 | 50% | 79% | ||
| 2.2 | 8 | 131 | 75% | 76% | ||
| 1.9 | 22 | 333 | 50% | 74% | ||
| 1 | 4 | 267 | 0% | 0% | ||
| Our list of github handles of Roche employees may not be complete. Our data on when employed (if they joined or left Roche) is manually added, so may not be exact. Generated with 23332 commits from 2014-06-12 to 2023-12-28. | ||||||
| 1 Person counted if ever employed by Roche, so contributions could occur while employed externally. | ||||||
| 2 Where a person joined or left Roche, commits are assigned to Roche only if employed when commit was made. | ||||||
| 3 Trend in % Roche comments month to month. X axis is not fixed across packages. | ||||||
This section looks at the lines of code per person present in the most recent version of the code. We exclude any code in the following folders in order to remove the risk of machine generated code skewing the results:
man/misc/inst/data/The number of people may differ, as in this section we are using the information from the persons .gitconfig file, rather than the Github user account information.
This data source uses information from the git commits, so we are relying on defining whether a person is Roche or not based on what email was in their .gitconfig file (or gh account if change made in github UI). This is currently very innacurate, so there are many false negatives where we miss Roche people contributing from a personal email account we do not know.
data_lines <- readRDS("cache/cleaned_lines.rds")
data_lines |>
mutate(
person = gsub(".*\\+","",person),
person = gsub("@users.noreply.github.com>","",person),
person = gsub("<","",person),
person = gsub(">","",person),
person = gsub("@github.com","",person),
# Is Roche
is_roche = case_when(
person %in% unique(roche_people$gh_username) ~ "Roche",
endsWith(person, '@roche.com') ~ "Roche",
endsWith(person, '@gene.com') ~ "Roche",
TRUE ~ "External"
),
Package = basename(repo)
) |>
filter(
!person %in% c("insights-engineering-bot","pharmaverse-bot")
) |>
left_join(
roche_pharmaverse |>
select(name, url_repo, url_hex, namespace),
by = c("Package" = "name")
) |>
group_by(namespace, repo, is_roche) |>
summarise(
lines = sum(lines), .groups = "drop"
) |>
pivot_wider(names_from = is_roche, values_from = lines) |>
mutate(
`% Roche` = case_when(
!is.na(Roche) ~ Roche/(Roche+External),
TRUE ~ 0
)
) |>
mutate(
Package = basename(repo)
) |>
select(
namespace, Package, `% Roche`
) |>
arrange(
desc(`% Roche`)
) |>
left_join(
roche_pharmaverse |>
select(name, url_repo, url_hex),
by = c("Package" = "name")
) |>
mutate(
Package = dplyr::case_when(
!is.na(url_hex) ~ glue::glue(
"<img src='{url_hex}' height='24'></img> <a href='{url_repo}'>`{Package}`</a>"),
TRUE ~ glue("<a href='{url_repo}'>`{Package}`</a>")
)
) |>
select(-c(url_repo, url_hex)) |>
group_by(namespace) |>
gt() |>
data_color(
columns = c(`% Roche`),
#rows = currency < 50,
method = "numeric",
palette = c("#F2D4FF","#007AC2"),
domain = c(0, 1)
) |>
tab_header(
title = "Lines of code attributed to Roche authors",
subtitle = md("Data comes from `git blame` of current package without `/docs`, `man` or `inst` folders")
) |>
tab_source_note(source_note = md(
glue("Our list of github handles of Roche employees may not be complete. Our data on when employed (if they joined or left Roche) is manually added, so may not be exact.")
)) |>
fmt_percent(
columns = c(`% Roche`),
decimals = 0
) |>
tab_footnote(
footnote = md("A line of code is attributed to Roche if the person *ever* worked for Roche. `xportr` is an example where lines of code were contributed before the person joined Roche's team."),
locations = cells_column_labels(columns = `% Roche`)
) |>
tab_options(
row_group.background.color = "#FAC9B5"
) |>
fmt_markdown(columns = c(Package))| Lines of code attributed to Roche authors | |
Data comes from git blame of current package without /docs, man or inst folders |
|
| Package | % Roche1 |
|---|---|
| pharmaverse | |
| 100% | |
| 100% | |
| 98% | |
| 93% | |
| 93% | |
| 89% | |
| 87% | |
| 83% | |
| 75% | |
| 68% | |
| 60% | |
| 57% | |
| 55% | |
| 50% | |
| 48% | |
| 47% | |
| 17% | |
| 0% | |
| 0% | |
| 0% | |
| openstatsware | |
| 100% | |
| 98% | |
| 90% | |
| 64% | |
| 63% | |
| 0% | |
| Our list of github handles of Roche employees may not be complete. Our data on when employed (if they joined or left Roche) is manually added, so may not be exact. | |
1 A line of code is attributed to Roche if the person ever worked for Roche. xportr is an example where lines of code were contributed before the person joined Roche’s team. |
|
Issues, and comments on issues, incorporates an additional form of collaboration that does not require writing code.
In [5]:
n_issues_pharmaverse <- issues_cleaned |>
filter(namespace == "pharmaverse") |>
dplyr::pull(url) |> n_distinct()
n_issues_openstatsware <- issues_cleaned |>
filter(namespace == "openstatsware") |>
dplyr::pull(url) |> n_distinct()
n_comments_pharmaverse <- issues_cleaned |>
filter(namespace == "pharmaverse") |>
nrow()
n_comments_openstatsware <- issues_cleaned |>
filter(namespace == "openstatsware") |>
nrow()There are 1808 issues and 6165 comments in the pharmaverse. There are 205 issues and 1530 in open statsware.
issues_cleaned |> ungroup() |>
left_join(
roche_people |>
select(gh_username, `Roche start`, `Roche end`) |> unique(),
by = c("author" = "gh_username")
) |>
mutate(
Month = as.Date(paste0(format(as.Date(comment_created), "%Y-%m"),"-01")),
# No set if commit is Roche or not
is_roche = case_when(
is.na(`Roche start`) &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
created > `Roche start` & created < `Roche end` &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
TRUE ~ "External"
)
) |>
group_by(namespace, Month) |>
summarise(
commits_roche = sum(is_roche == "Roche"),
commits_all = n()
) |> ungroup() |>
filter(commits_all >= 10) |>
mutate(
`% comments from Roche` = commits_roche/commits_all
) |>
ggplot(aes(x=Month, y=`% comments from Roche`)) +
geom_line(alpha = 0.3)+
geom_point(aes(size = commits_all), alpha = 0.5) +
ggthemes::theme_hc() +
facet_grid(rows = vars(namespace)) +
scale_y_continuous(labels = scales::percent_format(accuracy = 1)) +
labs(
title = "Monthly % of issue comments from Roche",
subtitle = "Months with <10 total comments excluded",
size = "Total comments"
)`summarise()` has grouped output by 'namespace'. You can override using the
`.groups` argument.
issues_cleaned |> ungroup() |>
group_by(url) |>
slice(1) |>
left_join(
roche_people |>
select(gh_username, `Roche start`, `Roche end`) |> unique(),
by = c("author" = "gh_username")
) |>
mutate(
Month = as.Date(paste0(format(as.Date(comment_created), "%Y-%m"),"-01")),
# No set if commit is Roche or not
is_roche = case_when(
is.na(`Roche start`) &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
created > `Roche start` & created < `Roche end` &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
TRUE ~ "External"
)
) |>
group_by(namespace, Month) |>
summarise(
commits_roche = sum(is_roche == "Roche"),
commits_all = n()
) |> ungroup() |>
filter(commits_all >= 10) |>
mutate(
`% issues from Roche` = commits_roche/commits_all
) |>
ggplot(aes(x=Month, y=`% issues from Roche`)) +
geom_line(alpha = 0.3)+
geom_point(aes(size = commits_all), alpha = 0.5) +
ggthemes::theme_hc() +
facet_grid(rows = vars(namespace)) +
scale_y_continuous(labels = scales::percent_format(accuracy = 1)) +
labs(
title = "Monthly % of issues opened by Roche",
subtitle = "Months with <10 total issues excluded",
size = "Total issues"
)`summarise()` has grouped output by 'namespace'. You can override using the
`.groups` argument.
left_join(
# Issues data ------------------------------------------------------------
issues_cleaned |>
left_join(
roche_people |>
select(gh_username, `Roche start`, `Roche end`) |> unique(),
by = c("author" = "gh_username")
) |>
mutate(
created = as.Date(created),
# No set if commit is Roche or not
is_roche = case_when(
is.na(`Roche start`) &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
created > `Roche start` & created < `Roche end` &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
TRUE ~ "External"
)
) |>
group_by(Package,issue_number) |>
slice(1) |>
group_by(Package) |>
summarise(
Issues = n(),
Roche = sum(is_roche == "Roche"),
`% Issues` = Roche/Issues,
.groups = "drop"
) |>
select(
Package, Issues, `% Issues`
),
# Comments data ------------------------------------------------------------
issues_cleaned |>
left_join(
roche_people |>
select(gh_username, `Roche start`, `Roche end`) |> unique(),
by = c("author" = "gh_username")
) |>
mutate(
created = as.Date(comment_created),
# No set if commit is Roche or not
is_roche = case_when(
is.na(`Roche start`) &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
created > `Roche start` & created < `Roche end` &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
TRUE ~ "External"
)
) |>
group_by(Package) |>
summarise(
Comments = n(),
Roche = sum(is_roche == "Roche"),
`% Comments` = Roche/Comments,
.groups = "drop"
) |>
select(
Package, Comments, `% Comments`
),
by = "Package"
) |> ## End of making two datasets
# Join with pharmaverse data
left_join(
roche_pharmaverse |>
select(name, url_repo, url_hex, namespace),
by = c("Package" = "name")
) |>
group_by(namespace) |>
# Make a trend for Comments
left_join(
issues_cleaned |>
left_join(
roche_people |>
select(gh_username, `Roche start`, `Roche end`) |> unique(),
by = c("author" = "gh_username")
) |>
mutate(
created = as.Date(comment_created),
Monthly = format(as.Date(comment_created), "%Y-%m"),
# No set if commit is Roche or not
is_roche = case_when(
is.na(`Roche start`) &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
created > `Roche start` & created < `Roche end` &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
TRUE ~ "External"
)
) |>
group_by(full_name, Monthly) |>
summarise(
`Roche commits` = sum(is_roche == "Roche"),
Commits = n()
) %>%
mutate(
Package = basename(full_name),
`% Comments` = `Roche commits`/Commits
) |>
arrange(Monthly) %>%
summarise(
`Comments trend` = list(`% Comments`), .groups = "drop"
) |>
ungroup() |>
mutate(
Package = basename(full_name)
) |>
select(
Package, `Comments trend`
),
by = "Package"
) |>
# Make a trend for issues
left_join(
issues_cleaned |>
group_by(url) |> slice(1) |>
left_join(
roche_people |>
select(gh_username, `Roche start`, `Roche end`) |> unique(),
by = c("author" = "gh_username")
) |>
mutate(
created = as.Date(comment_created),
Monthly = format(as.Date(comment_created), "%Y-%m"),
# No set if commit is Roche or not
is_roche = case_when(
is.na(`Roche start`) &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
created > `Roche start` & created < `Roche end` &
tolower(author) %in% tolower(roche_people$gh_username) ~ "Roche",
TRUE ~ "External"
)
) |>
group_by(full_name, Monthly) |>
summarise(
`Roche commits` = sum(is_roche == "Roche"),
Commits = n()
) %>%
mutate(
Package = basename(full_name),
`% Issues` = `Roche commits`/Commits
) |>
arrange(Monthly) %>%
summarise(
`Issues trend` = list(`% Issues`), .groups = "drop"
) |>
ungroup() |>
mutate(
Package = basename(full_name)
) |>
select(
Package, `Issues trend`
),
by = "Package"
) |>
# Format Package
mutate(
Package = dplyr::case_when(
!is.na(url_hex) ~ glue::glue(
"<img src='{url_hex}' height='24'></img> <a href='{url_repo}'>`{Package}`</a>"),
TRUE ~ glue("<a href='{url_repo}'>`{Package}`</a>")
)
) |>
select(-c(url_repo, url_hex)) |>
select(namespace, Package, Issues, Comments,
`% Issues`,`Issues trend`, `% Comments`,
`Comments trend`) |>
gt() |>
data_color(
columns = c(`% Issues`,`% Comments`),
#rows = currency < 50,
method = "numeric",
palette = c("#F2D4FF","#007AC2"),
domain = c(0, 1)
) |>
tab_header(
title = "Issue and Comments from Roche vs External collaborators",
subtitle = md("Data is captured going back 12 months only.")
) |>
tab_source_note(source_note = md(
glue("Our list of github handles of Roche employees may not be complete. Our data on when employed (if they joined or left Roche) is manually added, so may not be exact.")
)) |>
fmt_percent(
columns = c(`% Issues`,`% Comments`),
decimals = 0
) |>
tab_options(
row_group.background.color = "#FAC9B5"
) |>
tab_spanner(
label = "Roche + Externals",
columns = c(Issues, Comments)
) |>
tab_spanner(
label = "Roche",
columns = c(`% Issues`,`Issues trend`, `% Comments`, `Comments trend`)
) |>
fmt_markdown(columns = c(Package)) |>
gt_plt_sparkline(
`Comments trend`,
label = FALSE, same_limit = TRUE,
palette = c("black", rep("transparent", 4))
) |>
gt_plt_sparkline(
`Issues trend`,
label = FALSE, same_limit = TRUE,
palette = c("black", rep("transparent", 4))
) |>
tab_footnote(
footnote = md("Trend in % coming from Roche over last 12 months."),
locations = cells_column_labels(columns = c(`Issues trend`, `Comments trend`))
)`summarise()` has grouped output by 'full_name'. You can override using the
`.groups` argument.
`summarise()` has grouped output by 'full_name'. You can override using the
`.groups` argument.
`geom_line()`: Each group consists of only one observation. ℹ Do you need to
adjust the group aesthetic?
`geom_line()`: Each group consists of only one observation. ℹ Do you need to
adjust the group aesthetic?
`geom_line()`: Each group consists of only one observation. ℹ Do you need to
adjust the group aesthetic?| Issue and Comments from Roche vs External collaborators | ||||||
| Data is captured going back 12 months only. | ||||||
| Package | Roche + Externals | Roche | ||||
|---|---|---|---|---|---|---|
| Issues | Comments | % Issues | Issues trend1 | % Comments | Comments trend1 | |
| pharmaverse | ||||||
| 415 | 1680 | 51% | 52% | |||
| 85 | 238 | 38% | 31% | |||
| 35 | 86 | 77% | 78% | |||
| 62 | 107 | 81% | 80% | |||
| 59 | 138 | 7% | 7% | |||
| 19 | 34 | 100% | 100% | |||
| 47 | 95 | 49% | 46% | |||
| 8 | 14 | 12% | 7% | |||
| 5 | 17 | 20% | 29% | |||
| 1 | 2 | 0% | 0% | |||
| 193 | 515 | 92% | 91% | |||
| 127 | 592 | 80% | 81% | |||
| 49 | 233 | 96% | 98% | |||
| 72 | 253 | 85% | 77% | |||
| 77 | 264 | 81% | 78% | |||
| 57 | 158 | 77% | 80% | |||
| 209 | 1016 | 85% | 79% | |||
| 187 | 385 | 97% | 95% | |||
| 101 | 338 | 15% | 24% | |||
| openstatsware | ||||||
| 44 | 214 | 0% | 0% | |||
| 153 | 388 | 100% | 100% | |||
| 76 | 244 | 100% | 100% | |||
| 128 | 580 | 55% | 54% | |||
| 10 | 37 | 60% | 35% | |||
| 22 | 67 | 86% | 94% | |||
| Our list of github handles of Roche employees may not be complete. Our data on when employed (if they joined or left Roche) is manually added, so may not be exact. | ||||||
| 1 Trend in % coming from Roche over last 12 months. | ||||||