CODEWITHSUNDEEP
r
Paul
Paul

Reputation: 696

Extract rowname of rows containing unique combination of TRUE FALSE

I am attempting to extract the rowname that marking rows that contain a single TRUE value. Please see the following example.

                  Col1  Col2  Col3
Npat             FALSE  TRUE FALSE
Ttc30a1           TRUE FALSE FALSE
Gga1              TRUE  TRUE  TRUE
Gga3             FALSE FALSE FALSE
Gga2             FALSE FALSE  TRUE

I would like to extract the rownames of the rows containing a single unique "TRUE" value and retain information about which column the true value was in. The desired output might look like:

                   Col1  Col2  Col3
Npat             FALSE  TRUE FALSE
Ttc30a1           TRUE FALSE FALSE
Gga2             FALSE FALSE  TRUE

I attempted to use grep and -grep but couldn't get it recognize multiple values in multiple columns. I am sure there is a simple way to do this that I haven't been able to track down yet.

Upvotes: 0

Views: 143

Answers (3)

hdkrgr
hdkrgr

Reputation: 1736

You can simply treat booleans as if they were numeric. In base R:

df[with(df, Col1 + Col2 + Col3  == 1) ,]

Upvotes: 2

Pierre Lapointe
Pierre Lapointe

Reputation: 16277

Or you could use rowSums(df)==1:

df1 <-read.table(text="Row  Col1  Col2  Col3
Npat             FALSE  TRUE FALSE
Ttc30a1           TRUE FALSE FALSE
Gga1              TRUE  TRUE  TRUE
Gga3             FALSE FALSE FALSE
Gga2             FALSE FALSE  TRUE",header=TRUE, row.names=1,stringsAsFactors=FALSE)

df1[rowSums(df1)==1,]

         Col1  Col2  Col3
Npat    FALSE  TRUE FALSE
Ttc30a1  TRUE FALSE FALSE
Gga2    FALSE FALSE  TRUE

Upvotes: 2

Andre Elrico
Andre Elrico

Reputation: 11480

somewhat similar approach: (more general solution)

use apply to sum row-wise.

df[apply(df,1,sum)==1,]

Upvotes: 1

Related Questions