R-在多个条件下进行向量化

Question

I am trying to calculate a heat stress index that depends on three distinct temperatures, following this expression:

My R implementation works OK for single temperature values:

# Set temperatures
Teff=34
Tcr=33
Tlim=40

# Apply conditions
if (Teff < Tcr) {
  hsa = 1
} else if (Teff >= Tcr & Teff < Tlim) {
  hsa = 1 - ((Teff - Tcr)/(Tlim - Tcr))
} else if (Teff >= Tlim) {
  hsa = 0
}

hsa
  [1] 0.8571429

However, if I try to calculate hsa for a range of Teff , like this:

Teff=seq(30,40,1)

I get the following warning:

Warning message:
In if (Teff < Tcr) { :
  the condition has length > 1 and only the first element will be used

Which apparently occurs because if() is not vectorized and therefore evaluates only the first element of the vector.

I learned about ifelse() , which is vectorized if() , but I'm not sure how it can be used for multiple conditions.

So, my question is: what would be an alternative, vectorized way to calculate my hsa index using vectors instead of scalars?

Answer 1

How about this?

It's the same function as above, but without the explicit if and else ?

> Teff=seq(30,40,1)
> hsa<- 1*(Teff<Tcr) + (1 - (Teff - Tcr)/(Tlim - Tcr))*(Teff >= Tcr & Teff < Tlim)
> hsa
 [1] 1.0000000 1.0000000 1.0000000 1.0000000 0.8571429 0.7142857 0.5714286 0.4285714 0.2857143 0.1428571 0.0000000

** Note that you can add + 0*(Teff>=Tlim) in the end, but it wouldn't change anything, because the assigned value would be 0 anyways.

If you really want to use ifelse , then you have to nest them, so it should be something like this:

> hsa<- ifelse(Teff<Tcr, 1,
               ifelse(Teff >= Tcr & Teff < Tlim, 
               (1 - (Teff - Tcr)/(Tlim - Tcr)), 0))
> hsa
     [1] 1.0000000 1.0000000 1.0000000 1.0000000 0.8571429 0.7142857 0.5714286 0.4285714 0.2857143 0.1428571 0.0000000

Answer 2

Correct way to think about this kind of problem is to think about "logical indexing".

Example

Answer 3

@kandel is right. The better way is to use logical indexing.

Define the logical condition

idx <- iris$Species == "setosa"
new_iris = iris[idx,]

In your example, the following

if (Teff < Tcr) {
  hsa = 1
} else if (Teff >= Tcr & Teff < Tlim) {
         hsa = 1 - ((Teff - Tcr)/(Tlim - Tcr))
       } else if (Teff >= Tlim) {
hsa = 0
}

would translate to

idx <- Teff< Tcr
hsa[idx] <- 1 # and so on so forth