“../”相对路径目录不起作用 R

Question

I often have to work with shared data/code from Dropbox, and all of the files are usually loaded in as df <- import("../data/branch/file_name.csv")

However, this never loads on my computer, so I always have to type in my full directory as df <- import("C:/Users/Name/Dropbox/Folder1/Folder2/data/branch/file_name.csv")

Which makes it difficult for others to reproduce the code. Is there any way to fix this so that R can read the "../" part properly? I always get this error: Error: path does not exist: ../data/branch/file_name.csv

(I used the import function as an example but this problem occurs with other packages or loading functions as well)

I've also tried using "~/" but the problem persists. The shared code I work with always uses "../" though so I'd prefer a solution that gets that to work instead so that I don't have to change the original script each time I need to use it.

Thank you very much.

Edit: Despite the conversation below, I am still having trouble with this. I tried setting the working directory and then using "../" and it sometimes works for read_xls() (it was working a few hours ago, not anymore, depsite doing the exact same thing and setting the same working directory) but not for rio::import() or other reading functions. I've even tried setting my working directory to the exact same folder and using "./" but no luck. Errors now say "Cannot open the connection" or "No such file"

Answer 1

You can try setting your working directory(wd) to the current folder first, then use read.csv to import the data.

1. To set the working directory, first obtain the current directory using rstudioapi::getActiveDocumentContext()$path then use setwd() to set the wd to this path.
2. Then use whatever function to import your data.

For example, let's say your R code is saved in the "branch" folder, then the following code should work just fine:

setwd(dirname(rstudioapi::getActiveDocumentContext()$path))
mydata <- read.csv("./file_name.csv")
head(mydata)

Otherwise, if your R code was saved in a subfolder under Folder2 called code such that your R script's directory is: "C:/Users/Name/Dropbox/Folder1/Folder2/code/mycode.R" , then the following should work:

setwd(dirname(rstudioapi::getActiveDocumentContext()$path))
mydata <- read.csv("../data/branch/file_name.csv")
head(mydata)

Answer 2

It looks like the person you are collaborating with has their working directory set to a separate folder within Folder2 (perhaps a folder with all the.R code?). The .. means that the relative path is referencing the next folder up from the working directory (the next folder up is presumably Folder2).

To solve your problem, every time you start working you need to set your working directory to the same folder that your collaborator is using for their working directory. For example, this may be something like setwd("C:/Users/Name/Dropbox/Folder1/Folder2/RCode") . Unfortunately you will have to do that every time you open up the script, but if you do it at the very beginning, then the rest of the code should work.

Working directories can be hard to manage, especially when collaborating. A better solution to the working directory problem is to create a project in R Studio and collaborate on that project using GitHub. If you don't have experience with GitHub, these videos are an approachable resource for helping you get started.