An alternative display for profiling information

profr provides an alternative data structure and visual rendering for the profiling information generated by Rprof.


profr provides an alternative data structure and display for profiling data. It still uses Rprof() to collect the data, but outputs a data.frame which should be easier to manipulate. It also implements a novel visualisation which allows you to see the time taken by each function, as well as the context in which it was called.

To get started, try:

install.packages("profr")
library(profr)
p <- profr(my.slow.function())
plot(p)

Two built in examples are:

plot(nesting_prof)
plot(reshape_prof)

(and the second has helped me to considerably speed up (5-20x) the development version of reshape)

News

Version 0.3.1

  • Fixes for outstanding R CMD check problems

Version 0.3

  • Rewritten parsing to be about 3x faster

  • Corrected bug in parsing that lead to incorrect collapsing of non-adjacent functions

  • Get ggplot2.profr working again.

  • Updated documentation
  • Better namespace support

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("profr")

0.3.1 by Hadley Wickham, 4 years ago


https://github.com/hadley/profr


Report a bug at https://github.com/hadley/profr/issues


Browse source code at https://github.com/cran/profr


Authors: Hadley Wickham <h.wickham@gmail.com>


Documentation:   PDF Manual  


Task views: High-Performance and Parallel Computing with R


MIT + file LICENSE license


Imports stringr, plyr

Suggests ggplot2


Suggested by bossMaps, icd9.


See at CRAN