How to create a new column with for and if/else

Hi guys,

I'd like to create a new column of a dataframe using for and if/else functions. My dataframe ds1 has the following 4 columns:

Sample_Name     Sample_Well     Sentrix_ID     Sentrix_Position
pool_women      A01             2,04426E+11    R01C01
213141          B01             2,04426E+11    R02C0
pool_men        C01             2,04426E+11    R03C0
253141          D01             2,04426E+11    R04C01
202196          E01             2,04426E+11    R05C01
200569          F01             2,04426E+11    R06C01
242196          G01             2,04426E+11    R07C01

Now, I want to create a new column Sample_group, in which I should find 0 for samples starting with "21" and "20" in Sample_Name, 1 for samples starting with "25" and "24" in Sample_Names and 2 for the others (pool_women and pool_men), as following:

Sample_Name     Sample_Well     Sentrix_ID     Sentrix_Position     Sample_group
pool_women      A01             2,04426E+11    R01C01               2
213141          B01             2,04426E+11    R02C01               0
pool_men        C01             2,04426E+11    R03C01               2
253141          D01             2,04426E+11    R04C01               1
202196          E01             2,04426E+11    R05C01               0
200569          F01             2,04426E+11    R06C01               0
242196          G01             2,04426E+11    R07C01               1

I wrote the following code:

variables <- colnames(ds1[,which(colnames(ds1)=="Sample_Name")])

for(i in variables) {
  if(gsub("(^\\d{2}).*", "\\1", i) == "21" | gsub("(^\\d{2}).*", "\\1", i) == "20") {ds1$Sample_group1 <- 0}
  if(gsub("(^\\d{2}).*", "\\1", i) == "24" | gsub("(^\\d{2}).*", "\\1", i) == "25") {ds1$Sample_group1 <- 1}
  else {ds1$Sample_group1 <- 2}
}

However, I found only 1 at the Sample_group column for all samples.

What's wrong with my code?

Thank you!

• 1,222 views • link

updated 3.1 years ago by Ram

• written 5.2 years ago by Fede_Santos95 •

input:

> df

  Sample_Name
1  pool_women
2      213141
3    pool_men
4      253141
5      202196
6      200569
7      242196

output:

> df$group = with (df, 
+            ifelse (grepl("^20|^21", Sample_Name),0, 
+            ifelse(grepl("^25|^24",  Sample_Name),1,2 )))

> df
  Sample_Name group
1  pool_women     2
2      213141     0
3    pool_men     2
4      253141     1
5      202196     0
6      200569     0
7      242196     1

with dplyr:

df %>%
    mutate(across(
        .cols = Sample_Name,
        ~ ifelse (grepl("^20|^21", .),0, ifelse(grepl("^25|^24", .),1,2 )),
        .names = "group"
    ))

• 0 views • link

written 5.2 years ago by cpad0112

1 answer

Don't use for loops, that will access every single row, which will take ages if you have large data.frames.

#/ Example data:
df <- data.frame(Sample_Name=c("200", "211", 
                               "240", "251", 
                               "345", "456"))

#----------------------------------------------------------
# BASE R:
#----------------------------------------------------------
#/ new column with "2" (so other):
df$Sample_Group <- rep("2", nrow(df))

#/ and now replace 20/21s with 0 and 24/25s with 1:
df$Sample_Group[grep("^20|^21", df$Sample_Name)] <- 0
df$Sample_Group[grep("^24|^25", df$Sample_Name)] <- 1

> df
Sample_Name Sample_Group
1         200            0
2         211            0
3         240            1
4         251            1
5         345            2
6         456            2

#----------------------------------------------------------
# TIDYVERSE
#----------------------------------------------------------
library(tidyverse)
df %>%
  mutate(Sample_Group = 
           case_when(str_detect(Sample_Name, "^20|^21") ~ "0",
                     str_detect(Sample_Name, "^24|^25") ~ "1",
                     !str_detect(Sample_Name, "^20|^21|^24|^25") ~ "2")
)

Sample_Name Sample_Group
1         200            0
2         211            0
3         240            1
4         251            1
5         345            2
6         456            2

The ^ means "starts with". The first solution is with only base R functions, the second one uses the tidyverse packages.

Next time please provide example data, e.g. via dput. If you have a data.frame named df you can run dput(df) and it will print an ASCII representation of the data which you can provide. That makes it easy to copy/paste your data rather than typing down things. Also try to provide a small but representative selection of the data or dummy data to keep the post short and readable.

• 0 views • link

written 5.2 years ago by ATpoint