The goal of metapip is to provide the user with a set of functions that allow him/her to work efficiently with all the PIP R packages. The technical world of PIP consist of several packages that interact to each to load, format, modify, and estimate data for the PIP system. Given that the number of packages has increased over time, it was necessary to create a meta package whose only objective is the proper management of all the other PIP packages.
metapip is highly inspired by the {tidyverse}, so that some of its functions are just a simple refactoring of the tidyverse functions. All the credit, then, goes to the Tidyerse team. Thank you!
In addition to the basic functionality of attaching and installing PIP-related R packages, metapip also works as the manager of those packages in your working environment such that you are always working with the right version of any of the PIP packages.
Installation
You can install the development version of metapip from GitHub with:
# install.packages("devtools")
devtools::install_github("PIP-Technical-Team/metapip")Setting up default branch in metapip
In metapip there are different ways to set up default branches. By default it uses options(metapip.default_branch) to provide a default branch value to all packages. When you load the package it is set to “DEV_v2”. You can override that default setting on per package basis by updating options with custom default branch like options(metapip.custom_branch = list(pipapi_branch = "DEV", pipfaker_branch = "main")). So in this case, the default value for pipapi branch is “DEV” whereas the one for pipfaker is set to “main”. For rest of the pip core packages it remains as “DEV_v2”.
Functions and it’s usage in metapip
Since metapip works with Github API, you need to make sure that your GitHub credentials are properly set in you R environment. The best way to do so is using the gitcreds package. You may find a simple explanation about how you can generate a token for yourself and set it up here.
You can check if your credentials are properly set by running the following command.
check_github_token() |>
print()Let us now go through the functions in metapip.
-
This function as the name suggests gives you the available branches for a particular pip package on GitHub. It will display in the console the available branches.
get_branches("pipr")
#── These are available branches for pipr package:
#• 29-website-formatting
#• 61_global_end
#• DEV
#• add_retry
#• default_options
#• fix_popshare
#• gh-pages
#• httr2
#• main
#• testAt the same time it also returns the branch name as character vector invisibly.
out <- get_branches("pipr")
#── These are available branches for pipr package:
#• 29-website-formatting
#• 61_global_end
#• DEV
#• add_retry
#• default_options
#• fix_popshare
#• gh-pages
#• httr2
#• main
#• test
out
#[1] "29-website-formatting" "61_global_end" "DEV" "add_retry"
#[5] "default_options" "fix_popshare" "gh-pages" "httr2"
#[9] "main" "test" There is an argument called display which is set as TRUE by default. This argument controls if you want to show the branch names in the console or not.
out <- get_branches("pipr", display = FALSE)
out
#[1] "29-website-formatting" "61_global_end" "DEV" "add_retry"
#[5] "default_options" "fix_popshare" "gh-pages" "httr2"
#[9] "main" "test" -
This installs a specific branch for a particular package.
install_branch(package = "wbpip", branch = "DEV")The above command will install DEV branch of wbpip package.
This function will give you information of a particular branch in a package. For example, when was last commit done on this branch and by whom.
get_branch_info(package = "pipr", branch = "DEV")
# package branch_name last_commit_author_name last_update_time
#1 pipr DEV Tony Fujs 2023-09-04T13:06:47ZYou may pass more than one branch names.
get_branch_info(package = "wbpip", branch = c("PROD", "QA"))
# package branch_name last_commit_author_name last_update_time
#1 wbpip PROD Tony Fujs 2023-09-20T08:02:24Z
#2 wbpip QA tonyfujs 2023-07-14T08:50:33ZThis function installs the packages from the branch which was last updated i.e the latest branch.
If you pass no arguments to it, it will install latest branch for all pip core R packages. However, you can also restrict it to only a few packages as well.
install_latest_branch()
install_latest_branch(c("pipfun", "pipapi"))Get metadata information about pip packages.
out <- core_metadata()
out <- core_metadata(c("pipapi", "wbpip"))Install one or more core packages from a specific branch. If you have fresh installation of R then running install_pip_packages() would set you up with all pip packages.
install_pip_packages(branch = "test")
install_pip_packages(c("pipapi", "wbpip"), "test")This function checks the default branches for each packages and ensures that they are up-to-date. If they are not, it provides an option to update them with a single stroke.
init_metapip()
! The following packages do not have the updated version of default branch
installed: pipapi, pipload, pipfun, and pipaux
Do you want to install them now?
1: Yes
2: No
Selection: 1
── Attaching packages ──────────────────────────────────────────────────────────── metapip 0.0.0.9014 ──
✔ pipapi 1.3.19 (DEV)
✔ pipload 0.2.1 (DEV_v2)
✔ wbpip 0.1.6 (DEV)
✔ pipfun 0.3.0 (DEV_v2)
✔ pipdata 0.0.0.9017 (DEV)
✔ pipr 1.1.0 (main)
✔ pipster 0.0.1.9003 (DEV)
✔ pipaux 0.3.0 (DEV_v2)
✔ pipfaker 0.1.0 (main)An easier way to set options() for pip packages custom default branch.
set_custom_branch(pipr = 'main', 'pipapi' = 'DEV_v3')This is functionally equivalent to options(metapip.custom_branch = list(pipr_branch = "main", pipapi_branch = "DEV_v3")).