CODEWITHSUNDEEP
rcsv
Maximilian
Maximilian

Reputation: 4229

Read *.csv as list from multiple folders

I have thousands of *csv files in multiple folders ../t1/*.csv,../t2/*.csv,../t3/*.csv...etc.

I can upload the files from multiple folders as following:

filenames <- list.files(c("C:/Example/t1","C:/Example/t2"), pattern="*.csv", full.names=TRUE)
list.df <- lapply(filenames, read.csv)

However I have to type in all the directories C:/Example/t1, C:/Example/t2 etc. How to read all data (as list of data.frames) with one main directory somethign like: C:/Example/*?

Upvotes: 0

Views: 479

Answers (1)

Spacedman
Spacedman

Reputation: 94192

Using list.files with recursive=TRUE will search all folders under the first argument for matching files:

> list.files("./",recursive=TRUE)
[1] "a/a1.csv"    "a/a2.csv"    "a/notme.txt" "b/b1.csv"    "d/e/e1.csv"

That's all the files under my current directory, if I only want CSVs:

> list.files("./",recursive=TRUE,pattern="*.csv")
[1] "a/a1.csv"   "a/a2.csv"   "b/b1.csv"   "d/e/e1.csv"

Notice how it looks in the second-level d/e/ folder?

If you only want to go to a single, specific depth, try Sys.glob - these patterns match folders and files and these examples work from the current directory:

Only first level:

> Sys.glob("*/*.csv")
[1] "a/a1.csv" "a/a2.csv" "b/b1.csv"

Only second level:

> Sys.glob("*/*/*.csv")
[1] "d/e/e1.csv"

Upvotes: 4

Related Questions