When we switch from a *nix development environment to a Windows one, or reverse the process, we have to be confront with the dependancies problem.
In that, despite of the previous wheel packrat existing, which is built by Rstudio, a robust tool for virtualizing the environment, I am glad to make another wheel, an import function, which inspired from Python.
This import function will load the dependancies, and install these uninstalled one. These will improve the environment problem.
import <- function(packages){
for (i in 1:length(packages)){
package = packages[i];
char_package = package;
# char_package <- as.character(substitute(char_package))
loaded <- paste("package", char_package, sep = ":") %in% search() #
# is it existing the package
if(loaded){
library(char_package,character.only = TRUE)
cat(paste0("load ",char_package," package"),"\n")
}
else{
cat(paste0("can not find ",char_package,", we need to install it"),"\n");
install.packages(package);
library(char_package,character.only = TRUE)
}
}
}
# import dependencies
dependencies = c("RMySQL","ggplot2","httr")
import(dependencies)## can not find RMySQL, we need to install it## Warning: unable to access index for repository https://
## mran.revolutionanalytics.com/snapshot/2015-08-27/src/contrib## Warning: package 'RMySQL' is not available (for R version 3.2.2)## Loading required package: DBI
## Loading required package: methods## can not find ggplot2, we need to install it
##
## The downloaded binary packages are in
## /var/folders/f2/9jwh0h8s4y70r1jl3s7cq_5c0000gn/T//Rtmpqbv9uE/downloaded_packages
## can not find httr, we need to install it
##
## The downloaded binary packages are in
## /var/folders/f2/9jwh0h8s4y70r1jl3s7cq_5c0000gn/T//Rtmpqbv9uE/downloaded_packages