This article illustrates how to setup a working space to use the
mdJSONdictio functions, including how to install R, RStudio, and the
mdJSONdictio package.
What is R?
“R is a language and
environment for statistical computing and graphics.” - R Core Team
Team
Install R and RStudio on a Windows PC
Note: These instructions are catered to U.S. Fish & Wildlife
employees, but may be adapted to those with similar operating systems.
Visit the R and RStudio
download pages to find desktop installation instructions for Windows,
Mac, and Linux.
Get to know RStudio
There are four main panes in RStudio: Source (upper left), Console (lower left), Environment, History, Connections, Tutorial (upper right), and Files, Plots, Packages, Help, Viewer (lower right). Expand the Source pane with Ctrl + 1 or by clicking the double window icon in the upper right corner of the Console pane.
RStudio window
The Console pane is where code is executed. Code can be submitted
directly to the Console pane or written in “scripts” in the Source pane
and sent to the Console pane. Code written in the Source pane can be
saved for later, whereas the Console pane only keeps a short-term record
of executed code.
Operation tips
- Code in the Source pane is sent to the Console with Ctrl + Enter or by clicking the Run button.
- The blinking cursor in the Source pane indicates the line of code that will be executed. Chunks of code can be executed by highlighting multiple lines of code.
- Lines of code that start with # are “comments” and are not executed in the Console pane.
- A blue > at the bottom of the Console pane
indicates R finished processing the code.
For more information on how to use R and RStudio, visit the Tutorial
tab in the Environment, History, Connections, Tutorial pane, or refer to
the Hands-On
Programming with R website.
Install mdJSONdictio R package
Note: These instructions apply any computer system with R and RStudio.
Run installation code in RStudio
Paste the following in the RStudio Source pane and run each line of code. Run the code chunks separately to ensure each is successfully executed.
# Install devtools from CRAN
install.packages("devtools")
# Install mdJSONdictio from GitHub
devtools::install_github("hdvincelette/mdJSONdictio", ref = 'v0.2.2')Install or update dependent packages
mdJSONdictio requires other packages to work properly. R will automatically install any dependent packages not currently in your library during installation. R may also prompt you to update dependent packages already in your library. To update all packages, type “1” in the Console pane and hit Enter. If a pop-up asks “Do you want to install from sources the packages which need compilation?”, click Yes.
Verify mdJSONdictio installation
Run the following code to verify mdJSONdictio was successfully installed.
If R simply prints the executed code in the Console pane, then
installation was successful. If R returns the Error message “there is no
package called ‘mdJSONdictio’,” then installation failed. See Troubleshooting
Tips for more help.
Troubleshooting tips
Note: These instructions apply any computer system with R and
RStudio.
Package installation can be a fickle process, particularly when using GitHub. When a package fails to install, R returns Warning and Error messages which are often not descriptive or intuitive enough to understand the cause of failure. Most frequently, mdJSONdictio installation fails because dependent packages are not installed and updated, or command line settings are unsuitable. Solutions to these issues are described below. If you are not sure what the current problem is, work through each solution until mdJSONdictio installs successfully.
Failed installation due to dependent packages
Solution 1
Paste the following in the RStudio Source pane and run each line of code.
# Create list of dependent packages
packages<-c(
"purrr",
"readxl",
"rjson",
"stats",
"tibble",
"stringr",
"uuid",
"dplyr",
"plyr"
)
# Install/re-install packages
installed_packages <- packages %in% rownames(installed.packages())
if (any(installed_packages == FALSE)) {
install.packages(packages[!installed_packages])
}
# Check all packages installed
installed_packages <- packages %in% rownames(installed.packages())
if ((all(installed_packages == TRUE))) {
print("All dependent packages installed")
} else stop (
"One or more dependent packages did not install\n", toString(packages[installed_packages == FALSE]))If you see the the message “All dependent packages installed”, rerun
the mdJSONdictio
installation code. If you see the message “Error: One or more
dependent packages did not install”, install each of the listed packages
manually as described in Solution 2 below.
Solution 2
Uninstall the package in the Files, Plots, Packages, Help Viewer pane: Packages tab -> click the “x” next to the package. When a pop-up asks “Are you sure?”, click Yes.
Restarting RStudio session
Reinstall the package: Packages tab -> Install -> type in the package name -> Install
Reinstalling a package in RStudio
If you see the the message “cannot remove prior installation of package,” you will have to remove the package manually. Find the path to the package by running the following code:
# Get path to local R package library
.libPaths() Paste the file path in File Explorer, navigate to the package folder
and delete it. You may need to close R Studio in order to do so. Follow
the steps above to re-install the package.
Rerun
the mdJSONdictio
installation code.
Failed installation due to download method
Solution 1
Paste the following in the RStudio Source pane and run the line of code.
# Install packages using alternative method
options(download.file.method = "wininet")Note, “wininet” is a Windows only download method. If using a
different operating system, refer to https://www.rdocumentation.org/packages/utils/versions/3.6.2/topics/download.file
for alternative methods.
Rerun the mdJSONdictio
installation code.
Failed installation due to command line settings
Solution 1
Paste the following in the RStudio Source pane and run the line of code.
# Install mdJSONdictio with compiled code only for sub-architecture used by R CMD INSTALL
devtools::install_github("hdvincelette/mdJSONdictio", INSTALL_opts = c("--no-multiarch"))If mdJSONdictio installation fails again, install the package
manually as described in Solution 2 below.
Solution 2
Navigate to the mdJSONdictio GitHub repository and download the package zip: Code -> Download ZIP.
