New package in CRAN: PkgsFromFiles

Its been a while since I develop a CRAN package and this weekend I decided to work on a idea I had some time ago. The result is package PkgsFromFiles.

When working with different computers at home or work, one of the problems I have is installing missing packages across different computers. As an example, a script that works in my work computer may not work in my home computer. This is specially annoying when I have a fresh install of the operating system or R. In this case, I must manually install all packages, case by case. Instead of focusing on the script at hand, I spend considerable time finding and installing missing packages. When using laptops for teaching R, many times I had to wait for the installation of a package before continuing the class. With my new package, PkgsFromFiles, I can scan any folder of my computer and install all necessary packages before using them, as we will soon learn.

One of the available solutions to this problem is to use package pacman. It includes function p_load that will check if a package is available and, if not, install it from CRAN. However, for me, I like using library and require as it is consistent with my code format. Also, in a fresh R install, I rather install all my required packages in a single run so that I don’t have to wait later.

Package PkgsFromFiles solves this issue by finding and parsing all R related files (.R, .Rmd, .Rnw) from a given folder. It finds all calls to library() and require() and installs all packages that are not available locally.

Installation

# from cran (soon!)
install.packages('PkgsFromFiles')

# from github
if (!require(devtools)) install.packages('devtools')
devtools::install_github('msperlin/PkgsFromFiles')

Usage

The main function of the package is pff_find_and_install_pkgs, which will search and install missing packages from R files at a given directory. As an example, we’ll use my research folder from Dropbox. It contains all R scripts I have ever used in my research work. Let’s try it out:

# Evaluation is disable so it passes CRAN CHECKS, but you should be able to run it in your computer
library(PkgsFromFiles)

# target folder
my.dir <- '~/Dropbox/01-Pesquisa/'

df <- pff_find_and_install_pkgs(folder.in = my.dir)
## 
## Searching folder  ~/Dropbox/01-Pesquisa/
##  Found 34 files in 12 folders
##       R Scripts: 34 files
##       Rmarkdown files: 0 files
##       Sweave files: 0 files
## Checking available pkgs from https://cloud.r-project.org
## Checking and installing missing pkgs
## Installing dplyr Already installed
## Installing stringr   Already installed
## Installing GetDFPData    Already installed
## Installing xlsx  Already installed
## Installing googlesheets  Already installed
## Installing purrr Already installed
## Installing tidyverse Already installed
## Installing BatchGetSymbols   Already installed
## Installing lubridate Already installed
## Installing plm   Already installed
## Installing stargazer Already installed
## Installing RoogleVision  Installation failed, pkg not in CRAN
## Installing rvest Already installed
## Installing furrr Already installed
## Installing XML   Already installed
## Installing fst   Already installed
## Installing ggplot2   Already installed
## Installing genderBR  Already installed
## Installing sandwich  Already installed
## Installing lmtest    Already installed
## Installing MatchIt   Already installed
## Installing GetLattesData Already installed
## Installing RSelenium Already installed
## Installing httr  Already installed
## Installing GetTDData Already installed
## Installing rbcb  Already installed
## 
## Summary:
##  Found 25 packages already installed
##  Had to install 0 packages
##  Installation failed for 1 packages
##      1 due to package not being found in CRAN
##      0 due to missing dependencies or other problems
## 
## Check output dataframe for more details about failed packages

As you can see, function pff_find_and_install_pkgs will find all R related files recursively in the given folder. In this case, I have all packages locally so no installation was required. A summary in text is shown at the end of execution.

The output of the function is a dataframe with the details of the operation. Have a look:

dplyr::glimpse(df)
## Observations: 26
## Variables: 3
## $ pkg            <chr> "dplyr", "stringr", "GetDFPData", "xlsx", "goog...
## $ status.message <chr> "Already installed", "Already installed", "Alre...
## $ installation   <lgl> TRUE, TRUE, TRUE, TRUE, TRUE, TRUE, TRUE, TRUE,...

The package also includes function pff_find_R_files_from_folder, which will find all packages used in R related files from a given folder. It outputs a dataframe with several information about packages used in the found scripts.

df.files <- pff_find_R_files_from_folder(folder.in = my.dir)
## 
## Searching folder  ~/Dropbox/01-Pesquisa/
##  Found 34 files in 12 folders
##       R Scripts: 34 files
##       Rmarkdown files: 0 files
##       Sweave files: 0 files
dplyr::glimpse(df.files)
## Observations: 34
## Variables: 5
## $ files      <chr> "/home/msperlin/Dropbox/01-Pesquisa//01-Working Pap...
## $ file.names <chr> "01-Build_Presidents_Table.R", "02-DownloadPictures...
## $ extensions <chr> "R", "R", "R", "R", "R", "R", "R", "R", "R", "R", "...
## $ pkgs       <chr> "dplyr ; stringr ; GetDFPData ; xlsx", "googlesheet...
## $ n.pkgs     <int> 4, 4, 1, 3, 6, 6, 2, 4, 3, 6, 9, 6, 8, 5, 4, 1, 9, ...

I also wrote a simple function for plotting the most used packages for a given folder:

# target folder
my.dir <- '~/Dropbox/01-Pesquisa/'

# plot most used pkgs
p <- pff_plot_summary_pkgs(folder.in = my.dir)
## 
## Searching folder  ~/Dropbox/01-Pesquisa/
##  Found 34 files in 12 folders
##       R Scripts: 34 files
##       Rmarkdown files: 0 files
##       Sweave files: 0 files
print(p)

As you can see, I’m a big fan of the tidyverse!

Hope you guys find the package useful! Fell free to send any question to the comment section of the post or my email (marceloperlin@gmail.com).

Avatar
Marcelo S. Perlin
Associate Professor of Finance

Related

comments powered by Disqus