Search code examples
rdataframesubset

Subsetting data set based on logical test

I would like to subset a dataset based on a logical test.

My data look like:

A  B
1  2
3  4
5  7
2  1

Basically what I would like to do is to separate dataset into two sub-datasets, where one will contain all observations for which for a given row the reverse combination exists. So the desired output would look like:

data1
A B
1 2
2 1

And second:

data2
3 4
5 7

I know that subset() function allows for logical tests but I just don't know how to set up this one in R.

Solution

  • You can find the duplicated rows indices and then subset based on that:

    MySortedData <- data.frame(t(apply(df,1,sort)))
# X1 X2
# 1  1  2
# 2  3  4
# 3  5  7
# 4  1  2

MyDuplicates <- duplicated(MySortedData) | duplicated(MySortedData, fromLast=TRUE)
# [1]  TRUE FALSE FALSE  TRUE

MySubset2 <- df[!MyDuplicates,]
# A B
# 2 3 4
# 3 5 7

MySubset1 <- df[MyDuplicates,]
# A B
# 1 1 2
# 4 2 1