Read table by delimiter then by fixed width in R

问题

I have a tab-delimited file like this:

RS1->2001 HAPLO1 AAACAAGGAGGAGAAGGAAA ...
RS1->2001 HAPLO2 CAACAAAGAGGAGAAGGAAA ...
RS1->2002 HAPLO1 AAAAAAGGAGGAAAAGGAAA ...
RS1->20020 HAPLO2 CAACAAGGAGGAAGCAGAGC ...
RS1->20021 HAPLO2 CAACAAGGAGGAAGCAGAGC ...

In R we can easily read in these three columns, my problem is that I need separate the 3rd column character by character. The end result should be something like this:

RS1->2001 HAPLO1 A A A C  ...
RS1->2001 HAPLO2 C A A C  ...
RS1->2002 HAPLO1 A A A A  ...
RS1->20020 HAPLO2 C A A C  ...
RS1->20021 HAPLO2 C A A C  ...

I can first read the 3 columns in, then split each entry of the 3rd column into characters, but this is annoying, I would very much prefer to get it right from the start.

If the first two columns does not existe, I can achieve the goal with

read.fwf('test.csv', widths=rep(1, 300))

I am thinking whether I can read in the first 2 columns in by using the tab delimiter and then read the 3rd column by fixed width.

回答1:

The two main options that come to mind are strsplit (as mentioned in the comments and in @Ricardo's answer) and read.fwf. read.fwf won't work directly with your data, but it can work on a column of data that has already been read in if you use the textConnection() function.

Here's a basic example:

## Create a tab-separated file named "test.txt" in your working directory
cat("2001\tHAPLO1\tAAACAAGGAGGAGAAGGAAA\n",
    "2001\tHAPLO2\tCAACAAAGAGGAGAAGGAAA\n",
    "2002\tHAPLO1\tAAAAAAGGAGGAAAAGGAAA\n",
    "20020\tHAPLO2\tCAACAAGGAGGAAGCAGAGC\n",
    "20021\tHAPLO2\tCAACAAGGAGGAAGCAGAGC\n", 
    file = "test.txt")

## Read it in with `read.delim`
mydata <- read.delim("test.txt", header = FALSE, stringsAsFactors = FALSE)

## Use `read.fwf` on the third column
## Replace "widths" with whatever the maximum width is for that column
## If max width is not known, you can use something like
##    `widths = rep(1, max(nchar(mydata$V3)))`
cbind(mydata[-3], 
      read.fwf(file = textConnection(mydata$V3), widths = rep(1, 20)))
#      V1     V2 V1 V2 V3 V4 V5 V6 V7 V8 V9 V10 V11 V12 V13 V14 V15 V16 V17 V18 V19 V20
# 1  2001 HAPLO1  A  A  A  C  A  A  G  G  A   G   G   A   G   A   A   G   G   A   A   A
# 2  2001 HAPLO2  C  A  A  C  A  A  A  G  A   G   G   A   G   A   A   G   G   A   A   A
# 3  2002 HAPLO1  A  A  A  A  A  A  G  G  A   G   G   A   A   A   A   G   G   A   A   A
# 4 20020 HAPLO2  C  A  A  C  A  A  G  G  A   G   G   A   A   G   C   A   G   A   G   C
# 5 20021 HAPLO2  C  A  A  C  A  A  G  G  A   G   G   A   A   G   C   A   G   A   G   C

Note: If you did not use stringsAsFactors = FALSE, you would have to change your file argument to:

file = textConnection(as.character(mydata$V3))

回答2:

As @Ananda alludes to in the comments, strsplit if asked to split on "" will split every letter.

fContents <- read.csv("/path/to/file.csv")

 # This will chop it up for you.
 strsplit(fContents[, 3], "")

In order to combine it, use cbind

cbind(fContents[, -3],
       do.call(rbind, strsplit(fContents[, 3], ""))
      )

# or if you'd like to keep the columns ordered (and there are more than 3):
cbind(fContents[, 1:2], 
      do.call(rbind, strsplit(fContents[, 3], "")),
      fContents[, 4:ncol(fContents)]
)

回答3:

import csv
file_read = csv.reader(open('/path/to/file.csv','r'),delimiter='\t')
file_write = csv.writer(open('/path/to/newfile.csv','w'),delimiter='\t')
for i in file_read:
    first=i[0]
    second=i[1]
    third = i[3]
    splitchar = [k for k in third]
    outputdata = [first,second,splitchar]
    file_write.writerow(outputdata)

来源：https://stackoverflow.com/questions/17976628/read-table-by-delimiter-then-by-fixed-width-in-r

标签

data-manipulation