Search code examples
rprofilingperformance-testing

Examine which row/function in R code takes the most time within function?

I have a set of functions that run within a wrapper:

wrapper_func <- function(x,y,z,.....) {

t <- foo1(x,y)
kuku <- foo2(t,z)
....
final_res <- foo20(t, kuku, ...)

return(final_res)

}

It runs slowly and I want to understand who is the bottleneck/troublemaker. Please advise which function can perform deeper analysis (benchmark?microbenchmark?...) that will show the drilldown - which row/function takes the most time/resources?

I have found out another option

and go to the Memory profiling with lineprof chapter.

What do you think?

Solution

  • You can use Rprof to profile your R code and find the performance bottlenecks; here's a short example

    tmp <- tempfile()
Rprof(tmp)
example(glm)
Rprof()
summaryRprof(tmp)

    A more extensive description can be found at this R-bloggers article.