Why rbind does not work in function loop in R

Question

I am calculating the percentage of the impact within multiple rasters. I have one raster with a sum of whole variables (SUM). Then I have 9 rasters that I want to loop and calculate the percentage of the impact (Variables). I am using the function because I will use different inputs. When the function stops I want to have a text file or whatever is possible to export. I tried using rbind() but it does not work it is always empty when the function is executed. Here is the code:

Var_Sum_Impact(EU_HS_45_2041_SUM.r[[1]], EU_HS_45_2041_Var_Medi_Steppic.r)

results = data.frame(col1=numeric(), stringsAsFactors=FALSE)

Var_Sum_Impact = function(SUM, Variables) {
  for (i in 1:9){
    Var = cellStats(Variables[[i]], stat='mean')
    Su = cellStats(SUM, stat='mean')
    Perc = Var/(Su/100)
    print(Perc)
    results = rbind(results, Perc)
    }
}

And here is the result I get from that print but "results" is empty (0 obs.). The calculation is correct:

[1] 9.418741
[1] 20.39071
[1] 12.96036
[1] 5.910591
[1] 4.998413
[1] 21.37485
[1] 6.379199
[1] 4.383655
[1] 14.18348

Thank you Vincent. I have added also return in the function I have got this (what seems correct but the results are still empty when I write View(results)): col1 1 9.418741 2 20.390715 3 12.960364 4 5.910591 5 4.998413 6 21.374847 7 6.379199 8 4.383655 9 14.183477 — Martin Mokroš
– Martin Mokroš, Commented Sep 7, 2018 at 8:42

forestfanjoe · Accepted Answer · 2018-09-07 09:23:34Z

rbind isn't working as you expect here because you are trying to bind a data frame to a vector. So you need to either make Perc a data frame (or make results a vector and join each iteration using c()). You also need to return results after your for loop.

I also think instantiating results should happen inside your function

I think this should work (but haven't run it):

Var_Sum_Impact = function(SUM, Variables) {
  results = data.frame(col1=numeric(), stringsAsFactors=FALSE)
  for (i in 1:9){
    Var = cellStats(Variables[[i]], stat='mean')
    Su = cellStats(SUM, stat='mean')
    Perc = data.frame(col1 = Var/(Su/100), stringsAsFactors = FALSE)
    print(Perc)
    results = rbind(results, Perc)
    }
    results # return results
}
## call this after your function, right?
Var_Sum_Impact(EU_HS_45_2041_SUM.r[[1]], EU_HS_45_2041_Var_Medi_Steppic.r)

I'd also look into using lapply over the Variables object:

results_l <- lapply(Variables, function(V) {
        Var = cellStats(V, stat='mean')
        Su = cellStats(SUM, stat='mean')
        data.frame(col1 = Var/(Su/100), stringsAsFactors = FALSE)
}) ## produces a list of data frames you can rbind together
results <- do.call(rbind, results_l)

Collectives™ on Stack Overflow

Why rbind does not work in function loop in R

1 Answer 1

Comments

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Related