Main menu
You are here
models/nightly/BootstrapParallelNightly.R (snowfall test) fails in svn versions 1502 onwards.
If I svn up to any version 1502 or later, and run the BootstrapParallelNightly.R script, it fails with:
> names(submodels) <- omxExtractNames(submodels)
> topModel@submodels <- submodels
>
> modelResults <- mxRun(topModel, silent=TRUE, suppressWarnings=TRUE)
Error in checkForRemoteErrors(val) :
2 nodes produced errors; first error: no slot of name ".estimation" for this object of class "MxModel"
I am running 64bit R: R 2.10.1 GUI 1.31 Leopard build 64-bit (5537).
The error occurs with either snow v. 1.84 or 1.70. Currently have the latter:
> installed.packages()["OpenMx", "Version"]
[1] "999.0.0-1448"
> installed.packages()["snowfall", "Version"]
[1] "1.70"
> installed.packages()["snow", "Version"]
[1] "0.3-3"
For more information, my macbook pro configuration is like this:
Model Name: MacBook Pro
Model Identifier: MacBookPro5,1
Processor Name: Intel Core 2 Duo
Processor Speed: 2.8 GHz
Number Of Processors: 1
Total Number Of Cores: 2
L2 Cache: 6 MB
Memory: 4 GB
Bus Speed: 1.07 GHz
Boot ROM Version: MBP51.007E.B05
To ensure that nothing else in R memory was messing things up, I changed BootstrapParallelNightly.R to have this line at the beginning:
rm(list=ls())
I also modified the number of cpus in the script to be:
sfInit(parallel = TRUE, cpus = 2)
consistent with the hardware.
About the only thing I have not tried is a reboot. mspiegel has not been able to reproduce the error with the build system (using either R 2.11.1 or 2.10.1).
> installed.packages()["OpenMx", "Version"]
[1] "999.0.0-1448"
> installed.packages()["snowfall", "Version"]
[1] "1.70"
> installed.packages()["snow", "Version"]
[1] "0.3-3"
- Log in or register to post comments
Printer-friendly version
Comments
#1
The ".estimation" slot was eliminated from the code repository. You must be running an older version of the code:
euterpe:R mspiegel$ grep ".estimation" ~/Documents/OpenMx/trunk/R/*.Ryields no results on the svn trunk.
A scorched earth approach would be to delete everything recursively in /Library/Frameworks/R.framework. That will delete all your installed versions of R. Also see if you have anything in $HOME/Library/R.
#2
It's a little difficult to see how I could be running an older version of the code (by this you mean OpenMx code?).
svn up -r 1502 seems entirely happy. Except that it breaks the parallel code compared to v. 1501.
Would the scorched earth approach delete all my other installed libraries? That would not be very cool. Is there a possibility that any other installed library is causing the problem? It would be good to figure this out in case anyone else encounters the same problem, and I rather doubt whether many users are running parallel scripts with independent=TRUE.
I would like to get this working but I would also like greater insight as to how on earth it looks for a .estimation slot if there is no such reference in the code.
#3
Ok. I moved the $HOME version:
mv $HOME/Library/R $HOME/Library/R.old
which cured the problem. Running fine under svn r 1514. tbh, I don't know how I ended up with a local R 2.10 and an OpenMx in it. Also puzzled as to how
svn up -r 1501; cd trunk; make build install
was causing different behavior if it was being pre-empted by the version in $HOME/R. So I guess it was using some of what was in $HOME and otherwise using the recently installed version. Which means I must have been running a blended version somehow.
Michael-Neales-Macbook-Pro-4% dir $HOME/Library/R
total 0
drwxr-xr-x 4 neale neale 136 Sep 24 17:15 ./
drwx------+ 79 neale neale 2686 Nov 12 10:32 ../
drwxr-xr-x 3 neale neale 102 Sep 24 17:15 2.10/
drwxr-xr-x 3 neale neale 102 Apr 9 2008 2.6/
Michael-Neales-Macbook-Pro-4% mv $HOME/Library/R $HOME/Library/R.old
Michael-Neales-Macbook-Pro-4% dir $HOME/Library/R.old
total 0
drwxr-xr-x 4 neale neale 136 Sep 24 17:15 ./
drwx------+ 79 neale neale 2686 Nov 22 11:12 ../
drwxr-xr-x 3 neale neale 102 Sep 24 17:15 2.10/
drwxr-xr-x 3 neale neale 102 Apr 9 2008 2.6/
Michael-Neales-Macbook-Pro-4% dir $HOME/Library/R.old/2.10
total 0
drwxr-xr-x 3 neale neale 102 Sep 24 17:15 ./
drwxr-xr-x 4 neale neale 136 Sep 24 17:15 ../
drwxr-xr-x 6 neale neale 204 Nov 22 09:46 library/
Michael-Neales-Macbook-Pro-4% dir $HOME/Library/R.old/2.10/library/
BayesMendel/ OpenMx/ R.css latticedl/
Michael-Neales-Macbook-Pro-4% dir $HOME/Library/R.old/2.10/library/OpenMx/
total 56
drwxr-xr-x 17 root neale 578 Nov 22 09:12 ./
drwxr-xr-x 6 neale neale 204 Nov 22 09:46 ../
-rw-r--r-- 1 root neale 1570 Nov 22 09:12 CITATION
-rw-r--r-- 1 root neale 2324 Nov 22 09:12 DESCRIPTION
-rw-r--r-- 1 root neale 2956 Nov 22 09:12 INDEX
-rw-r--r-- 1 root neale 11350 Nov 22 09:12 LICENSE
drwxr-xr-x 9 root neale 306 Nov 22 09:12 Meta/
-rw-r--r-- 1 root neale 674 Nov 22 09:12 NAMESPACE
drwxr-xr-x 5 root neale 170 Nov 22 09:12 R/
drwxr-xr-x 5 root neale 170 Nov 22 09:12 data/
drwxr-xr-x 58 root neale 1972 Nov 22 09:12 demo/
drwxr-xr-x 7 root neale 238 Nov 22 09:12 help/
drwxr-xr-x 3 root neale 102 Nov 22 09:12 html/
drwxr-xr-x 3 root neale 102 Nov 22 09:12 libs/
drwxr-xr-x 5 root neale 170 Nov 22 09:12 npsol/
drwxr-xr-x 20 root neale 680 Nov 22 09:12 tools/
drwxr-xr-x 2 root neale 68 Nov 22 09:12 unit-tests/
#4
Resolved - not a code bug but an unexpected interaction with a $HOME/Library installed version of OpenMx.
#5
Automatically closed -- issue fixed for 2 weeks with no activity.