This is the proposed workflow and some latter parts
are done with Matlab.
I assist is masses of Matlab jobs using compiled Matlab
and run thousands of jobs both in on center, on our campus
and out in to other resources in the GRID.
My problem is not how to run a compiled Matlab job, but
how to interface the Matlab portion perhaps invisibly
and perhaps not to running with compiled matlab code.
I know and will never be a freesurfer person. I am only
trying to automate more of it. Is there perhaps a single
script which matches up to calls to Matlab entry points?
processed in parallel for two subjects three time points example;
recon-all -autorecon1 -subjid 1001a
recon-all -autorecon1 -subjid 1001b
recon-all -autorecon1 -subjid 1001c
recon-all -autorecon1 -subjid 1002a
recon-all -autorecon1 -subjid 1002b
recon-all -autorecon1 -subjid 1002c
.
.
.
After completion of -autorecon1
recon-all -autorecon2 -subjid 1001a
recon-all -autorecon2 -subjid 1001b
recon-all -autorecon2 -subjid 1001c
recon-all -autorecon2 -subjid 1002a
recon-all -autorecon2 -subjid 1002b
recon-all -autorecon2 -subjid 1002c
.
.
.
After completion of -autorecon2
recon-all -autorecon3 -subjid 1001a
recon-all -autorecon3 -subjid 1001b
recon-all -autorecon3 -subjid 1001c
recon-all -autorecon3 -subjid 1002a
recon-all -autorecon3 -subjid 1002b
recon-all -autorecon3 -subjid 1002c
.
.
After completion of -autorecon3
OPTIONAL COMPUTATION OF LGI WHICH REQUIRES MATLAB ~3 hours
recon-all -s 1001a -localGI
recon-all -s 1001b -localGI
recon-all -s 1001c -localGI
recon-all -s 1002a -localGI
recon-all -s 1002b -localGI
recon-all -s 1002c -localGI
recon-all -base 1001base3 -tp 1001a -tp 1001b -tp 1001c -all
recon-all -base 1002base3 -tp 1002a -tp 1002b -tp 1002c -all
.
.
.
After completion of -base runs
recon-all -long 1001a 1001base3 -all
recon-all -long 1001b 1001base3 -all
recon-all -long 1001c 1001base3 -all
recon-all -long 1002a 1002base3 -all
recon-all -long 1002b 1002base3 -all
recon-all -long 1001c 1002base3 -all
On 7/19/2011 2:03 PM, Bruce Fischl wrote:
> Hi Bill,
> what portions are you talking about? recon-all shouldn't use any matlab.
>
> cheers
> Bruce
> On
> Tue, 19 Jul 2011, Bill Taylor wrote:
>
>> So I peaked in archive and did not get a hit.
>>
>> The situation is that the University of Wisconsin
>> Madison site license prohibits us running Matlab
>> in a cluster unless it is compiled.
>>
>> I am assisting a Neurology lab with having a Condor
>> workflow which this time round will do both Matlab
>> and FSL post processing.
>>
>> To compile the Matlab one needs to compile the
>> entry points. But to run it, we'd have to change how it is called.
>>
>> Anyone have pointers for me?
>>
>> Bill Taylor
>> Center for High Throughput Computing
>> Condor Project
>> Computer Sciences
>> University of Wisconsin, Madison
>> 608-219-4430
>>
>> _______________________________________________
>> Freesurfer mailing list
>> Freesurfer@nmr.mgh.harvard.edu
>> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
>>
>>
>>
> _______________________________________________
> Freesurfer mailing list
> Freesurfer@nmr.mgh.harvard.edu
> https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
>
>
> The information in this e-mail is intended only for the person to whom it is
> addressed. If you believe this e-mail was sent to you in error and the e-mail
> contains patient information, please contact the Partners Compliance HelpLine
> at
> http://www.partners.org/complianceline . If the e-mail was sent to you in
> error
> but does not contain patient information, please contact the sender and
> properly
> dispose of the e-mail.
>
_______________________________________________
Freesurfer mailing list
Freesurfer@nmr.mgh.harvard.edu
https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
