How do I subset a data frame by row for ALL columns

Question

This seems like a very simple problem but I am unable to find a solution anywhere online. I want to subset a data frame to rows for which any column contains the string "string". This is how I want to subset the data:

subset <- df[which(df$V1 == 'string' | df$V2 == 'string' | df$V3=='string'),]

But my frame has many columns and it is very inefficient to list all of them. Is there a more efficient way to select rows for which ANY column contains 'string'?

Avoid names like `subset`. Not always important due to the presence of IDEs, but might spend too much time figuring out what went wrong if you ever need to use the function `subset`. Please also provide a sample of your data. — NelsonGon, Jun 26 '19 at 15:41
Similar post, just with logic of *dropping* those rows instead of *keeping* them: https://stackoverflow.com/q/44544596/5325862 — camille, Jun 26 '19 at 16:06

akrun · Accepted Answer · 2019-06-26T15:59:24.457

3

We can use rowSums to create a create a logical vector for subsetting the rows

nm1 <- paste0("V", 1:3)
df[rowSums(df[nm1] == "string") >  0,]

edited Jun 26 '19 at 15:59

answered Jun 26 '19 at 15:41

akrun

874,273
37
540
662

This is great. Thank you! – Harley McCaffery Jun 26 '19 at 15:58
1

Probably want `paste0`, not `patse0` – camille Jun 26 '19 at 15:59
1

Cool, sorry, I am a stack noob. It worked. – Harley McCaffery Jun 26 '19 at 16:02

How do I subset a data frame by row for ALL columns

1 Answers1