Yan Yan - 4 years ago 78
R Question

Read in a countinuous text file into data.frame

I have a text file that only has one column. It's like:


color 12
length 34
validity 90


color 15
length 20
validity 120


color 34
validity 79

There are 3 lines between samples, and 1 line between sample id and its attribute. Also, for sample3, the length record is missing.

I want to read this file into an R data.frame so that it looks like:

sample1 sample2 sample3
color 12 15 34
length 34 20 NA
validity 90 120 79

Answer Source

You've got a data cleaning problem. Here is my solution for you.

I copied and pasted your "TXT" file into a blank TextEdit document on Mac, and saved it as file.txt. The order as shown in your "TXT" file is required:

data <- unlist(read.table("file.txt", header=F, sep="\t", stringsAsFactors=F), use.names=F)

sample_names <- data[grep("sample", data), drop=T]
## [1] "sample1" "sample2" "sample3"

color <- data[grep("color", data), drop=T]
## "color 12" "color 15" "color 34"

length <- data[grep("length", data), drop=T]
length #note missing term, and requires manual coding
## [1] "length 34" "length 20"

length <- c(length, NA)
## [1] "length 34" "length 20" NA   

validity <- data[grep("validity", data), drop=T]
## [1] "validity 90"  "validity 120" "validity 79" 

## Assemble into data.frame:
assembled_df <- rbind(color, length, validity)
colnames(assembled_df) <- sample_names #update column names
##          sample1       sample2        sample3      
## color    "color 12"    "color 15"     "color 34"   
## length   "length 34"   "length 20"    NA           
## validity "validity 90" "validity 120" "validity 79"

Note that the code might not be generalizable. It is a matter of what the actual TXT file would look like. What's important is to learn to 1) know your data (which you do), 2) come up with a strategy, 3) and then a solution.

Recommended from our users: Dynamic Network Monitoring from WhatsUp Gold from IPSwitch. Free Download