@@ -37,13 +38,14 @@

Performance benefit by using compiled model definitions in mkin

Johannes Ranke

2016-10-06

2016-10-26

Benchmark for a model that can also be solved with Eigenvalues

This evaluation is taken from the example section of mkinfit. When using an mkin version equal to or greater than 0.9-36 and a C compiler (gcc) is available, you will see a message that the model is being compiled from autogenerated C code when defining a model using mkinmod. The mkinmod() function checks for presence of the gcc compiler using

Sys.which("gcc")

##            gcc 
@@ -73,25 +75,25 @@ mb.1 <- micr
 print(mb.1)

## Unit: milliseconds
 ##                   expr       min        lq      mean    median        uq
-##  deSolve, not compiled 6365.2773 6407.1609 6467.4083 6449.0446 6518.4738
-##       Eigenvalue based  872.0339  912.6754  926.7178  953.3169  954.0597
-##      deSolve, compiled  730.1252  740.7149  751.8881  751.3045  762.7695
+##  deSolve, not compiled 6298.7342 6308.6792 6343.9668 6318.6243 6366.5831
+##       Eigenvalue based  871.7379  880.7757  903.5267  889.8135  919.4211
+##      deSolve, compiled  724.9025  730.6729  732.9837  736.4432  737.0243
 ##        max neval cld
-##  6587.9031     3   b
-##   954.8026     3  a 
-##   774.2345     3  a

+## 6414.5420 3 c +## 949.0287 3 b +## 737.6054 3 a

autoplot(mb.1)

We see that using the compiled model is by a factor of 8.6 faster than using the R version with the default ode solver, and it is even faster than the Eigenvalue based solution implemented in R which does not need iterative solution of the ODEs:

rownames(smb.1) <- smb.1$expr
 smb.1["median"]/smb.1["deSolve, compiled", "median"]

##                         median
-## deSolve, not compiled 8.583796
-## Eigenvalue based      1.268882
+## deSolve, not compiled 8.579920
+## Eigenvalue based      1.208258
 ## deSolve, compiled     1.000000

Benchmark for a model that can not be solved with Eigenvalues

This evaluation is also taken from the example section of mkinfit.

FOMC_SFO <- mkinmod(
   parent = mkinsub("FOMC", "m1"),
@@ -108,28 +110,29 @@ smb.1["median"]/smbsmb.2 <- summary(mb.2)
 print(mb.2)


 ## Unit: seconds
-##                   expr       min        lq     mean    median        uq
-##  deSolve, not compiled 13.464194 13.498471 13.52712 13.532748 13.558577
-##      deSolve, compiled  1.328628  1.342168  1.35224  1.355707  1.364046
+##                   expr       min        lq      mean    median        uq
+##  deSolve, not compiled 13.265212 13.330161 13.412053 13.395109 13.485473
+##      deSolve, compiled  1.322466  1.326851  1.364827  1.331236  1.386007
 ##        max neval cld
-##  13.584407     3   b
-##   1.372385     3  a
+##  13.575837     3   b
+##   1.440779     3  a

smb.2["median"]/smb.2["deSolve, compiled", "median"]

##   median
 ## 1     NA
 ## 2     NA

autoplot(mb.2)

Here we get a performance benefit of a factor of 10 using the version of the differential equation model compiled from C code!

Here we get a performance benefit of a factor of 10.1 using the version of the differential equation model compiled from C code!

This vignette was built with mkin 0.9.44.9000 on

## R version 3.3.1 (2016-06-21)
 ## Platform: x86_64-pc-linux-gnu (64-bit)
 ## Running under: Debian GNU/Linux 8 (jessie)

## CPU model: Intel(R) Core(TM) i7-4710MQ CPU @ 2.50GHz