Replace symbols in R table

Question

[Inquisitive8995]

Hello,

I have a table in R which consists of 70k rows and 37 columns. Lot of cells have "./." which I want to modify and make it as "ab" . I tried to use gsub() but it does not give me the required output.

I used :

file <- gsub("./.","ab",file)

I want the change to happen throughout the file. Is there any other way with which I can modify it? Thanks in advance.

Input: Eg:

S.no chr pos gene_name S1 S2 S3
1        1    1290    X            ./.   1/1  ./.
2        1     5822    Y           0/1   ./.   ./.

Output

S.no chr pos gene_name S1 S2 S3
    1        1    1290    X            ab   1/1  ab
    2        1     5822    Y           0/1   ab   ab

It can be either ab or NA

Answer 1

[zx8754]

Try to use fixed match:

file <- gsub("./.","ab",file, fixed = TRUE)

Or

file[ file == "./." ] <- "ab"

---

Edit: Using example data provided by OP.

# example input data
df1 <- read.table(text = "
S.no chr pos gene_name S1 S2 S3
1        1    1290    X            ./.   1/1  ./.
2        1     5822    Y           0/1   ./.   ./.", header = TRUE, stringsAsFactors = FALSE)

df1
#   S.no chr  pos gene_name  S1  S2 S3
# 1    1   1 1290         X  ab 1/1 ab
# 2    2   1 5822         Y 0/1  ab ab

df1[, c("S1", "S2", "S3")][ df1[, c("S1", "S2", "S3")] == "./." ] <- "ab"
df1
#   S.no chr  pos gene_name  S1  S2 S3
# 1    1   1 1290         X  ab 1/1 ab
# 2    2   1 5822         Y 0/1  ab ab

[Inquisitive8995]

Thanks for your answer. The second one worked but it does not show "ab" instead it is blank and has "NA"s

[zx8754]

Provide reproducible example input data, and expected output.

[Inquisitive8995]

I have edited my question with an example input and output. Thanks

[zx8754]

Edited my answer, see if it works.

Answer 2

[cpad0112]

df1=read.csv("test.txt", header = T, strip.white = T, stringsAsFactors = F, sep = "\t")
library(stringr)
library(dplyr)
> df1 %>%  mutate_all(funs(str_replace_all(.,"\\.[/|\\|]\\.","ab")))
  S.no chr  pos gene_name  S1  S2 S3
1    1   1 1290         X  ab 1/1 ab
2    2   1 5822         Y 0/1  ab ab

You can also use apply function:

> library(stringr)
> apply(df1,2, function(x) str_replace_all(x,"\\.[/|\\|]\\.","ab"))
     S.no chr pos    gene_name S1    S2    S3  
[1,] "1"  "1" "1290" "X"       "ab"  "1/1" "ab"
[2,] "2"  "1" "5822" "Y"       "0/1" "ab"  "ab"

This is supposed to replace both ./. and .|.. test.txt is OP input text.

[Inquisitive8995]

I want only the ./. to be modified not the 1/1.

[cpad0112]

oops..typo. It doesn't replace any character other than . (./. or .|.). Edited OP. Inquisitive8995

[zx8754]

Moved your post to an answer, you might want to clean up your above comments, or edit them in into this post.

[cpad0112]

Thanks zx8754