Select rows when they contain certain string using R

Question

I have a data frame below:

Acct <- c(1001, 1002, 1003)
Tran <- c(01, 02, "ALL")
Group <- c(01, 01, 02)
DF1 <- data.frame(Acct, Tran, Group)

Now I need to select the rows where "ALL" shows up. The result should look like:

 Acct |  Tran  | Group
 1003 |   ALL  |  2

One thing to mention is that "ALL" can be in any column (not just "Tran") and there may be any number of columns.

Answer 1

You can use DF1 == "ALL" to get a logical matrix to find ALL. Then we can take the rowSums() of that matrix and keep the ones where the sum is not zero.

DF1[rowSums(DF1 == "ALL") != 0, ]
#   Acct Tran Group
# 3 1003  ALL     2

Answer 2

Use this line:

apply(DF1 == "ALL",1,any)

This will result in a T/F vector that corresponds to the rows that contain at least one "ALL"

DF1 == "ALL" gives:

> DF1 == "ALL"
      Acct  Tran Group
[1,] FALSE FALSE FALSE
[2,] FALSE FALSE FALSE
[3,] FALSE  TRUE FALSE

By using apply, as stated above, you get:

> apply(DF1 == "ALL",1,any)
[1] FALSE FALSE  TRUE

And if you want the rows:

> select <- apply(DF1 == "ALL",1,any)
> DF1[select,]

  Acct Tran Group
3 1003  ALL     2