From af7c6de4db9981ac814362c441fbac22c8faa2d7 Mon Sep 17 00:00:00 2001 From: Johannes Ranke Date: Thu, 24 Nov 2022 09:02:26 +0100 Subject: Start online docs of the development version --- docs/dev/articles/web_only/compiled_models.html | 25 ++++++++++++++----------- 1 file changed, 14 insertions(+), 11 deletions(-) (limited to 'docs/dev/articles/web_only/compiled_models.html') diff --git a/docs/dev/articles/web_only/compiled_models.html b/docs/dev/articles/web_only/compiled_models.html index ade86bc5..e9d80420 100644 --- a/docs/dev/articles/web_only/compiled_models.html +++ b/docs/dev/articles/web_only/compiled_models.html @@ -34,7 +34,7 @@ mkin - 1.2.0 + 1.2.2 @@ -78,7 +78,10 @@ Example evaluation of NAFTA SOP Attachment examples
  • - Some benchmark timings + Benchmark timings for mkin +
    • +
  • + Benchmark timings for saem.mmkin
    • @@ -109,7 +112,7 @@

    Performance benefit by using compiled model definitions in mkin

    Johannes Ranke

    -

    2022-11-01

    +

    2022-11-24

    Source: vignettes/web_only/compiled_models.rmd @@ -167,10 +170,10 @@ print("R package rbenchmark is not available") } 
    ##                    test replications relative elapsed
-## 4            analytical            1    1.000   0.186
-## 3     deSolve, compiled            1    1.656   0.308
-## 2      Eigenvalue based            1    2.102   0.391
-## 1 deSolve, not compiled            1   38.968   7.248
    +## 4 analytical 1 1.000 0.221 +## 3 deSolve, compiled 1 1.561 0.345 +## 2 Eigenvalue based 1 1.932 0.427 +## 1 deSolve, not compiled 1 33.629 7.432

    We see that using the compiled model is by more than a factor of 10 faster than using deSolve without compiled code.

    @@ -197,10 +200,10 @@ }
    ## Temporary DLL for differentials generated and loaded
    ##                    test replications relative elapsed
-## 2     deSolve, compiled            1    1.000   0.452
-## 1 deSolve, not compiled            1   29.431  13.303
    -

    Here we get a performance benefit of a factor of 29 using the version of the differential equation model compiled from C code!

    -

    This vignette was built with mkin 1.2.0 on

    +## 2 deSolve, compiled 1 1.000 0.482 +## 1 deSolve, not compiled 1 27.865 13.431 +

    Here we get a performance benefit of a factor of 28 using the version of the differential equation model compiled from C code!

    +

    This vignette was built with mkin 1.2.2 on

    ## R version 4.2.2 (2022-10-31)
 ## Platform: x86_64-pc-linux-gnu (64-bit)
 ## Running under: Debian GNU/Linux 11 (bullseye)
    -- cgit v1.2.1