any recommend for finding the intersection of two continuous variable in r

Question

How can I find the intersection percentages of some continuous variables, see the example below, please?

d1<-data.frame(Start=c(10, 8, 6, 4 ), End=c(14, 12, 9,17 ))

I want to check each row of the columns A and B overlap with the rest of the row, instead of for loop? For example,

d1[1,] %overlaps% d1[2,]

and d1[1,] %overlaps% d1[3,]and ..finally, d3[1,] %overlaps% d4[3,]

How to do that?

What's the desired output here? And what exactly do you mean by "overlaps"? Is that different from `%in%`? Are you just looking for duplicated rows? — MrFlick, Sep 03 '20 at 20:21
Ok, but how to compare each row with the other rows? For example, row 1 and row2, row1 and row3, row 1 and row 4, then row2 and row3.. — Melih Aras, Sep 03 '20 at 20:33
Possible duplicate https://stackoverflow.com/q/24480031/680068 — zx8754, Sep 03 '20 at 20:43

score 0 · Accepted Answer · answered Sep 03 '20 at 20:41

Something like the following determines if one segment in the real line defined by the endpoints Start and End overlaps another segment. The row combinations are created with combn and an anonymous function is applied to each combination.

`%overlaps%` <- function(X, Y){
  f <- function(x, y){
    a1 <- x[1] <= y[1] && y[1] <= x[2]
    a2 <- x[1] <= y[2] && y[2] <= x[2]
    a1 || a2
  }
  f(X, Y) || f(Y, X)
}

combn(1:nrow(d1), 2, function(x) {
  d1[x[1], ] %overlaps% d1[x[2], ]
})
#[1]  TRUE FALSE  TRUE  TRUE  TRUE  TRUE

score 0 · Answer 2 · answered Sep 03 '20 at 21:58

Note that Start in d1 has been given in a descending order, you only need to check if End values of next intervals are greater than the current Start value, e.g.,

> unlist(sapply(1:(nrow(d1)-1),function(k) d1$End[-(1:k)]>=d1$Start[k]))
[1]  TRUE FALSE  TRUE  TRUE  TRUE  TRUE

any recommend for finding the intersection of two continuous variable in r

2 Answers2