Hi Brian,

I haven't actually tried running it with -loadBvars and without -shcond 
yet. Unfortunately, I've left the lab for the day, but I'll let you know 
if that combination works within the next few days. And the FIRST patch 
will be greatly appreciated!

Thanks,

Zac


Brian Patenaude wrote:
> Hi,
>
> I believe this is an oversight on my part, we didn't include the
> conditional files corresponding to the new model (336 subjects vs 316
> subjects). Note that .bvars from the old models should not be used with
> new models. When generating the new .bvars, does the -loadBvars command
> work without the -shcond? There should be a patch to FIRST released very
> soon, I'll try and include the conditional matrices.
>
> Cheers,
>
> Brian
>
>
>
>
>   
>> Dear FSLers,
>>
>> I just recently upgraded to the latest version of FSL, and I have since
>> been
>> having issues using run_first with the -shcond and -loadBvars options.
>> Running run_first without these options seems to work fine, but each time
>> I
>> try to include them I receive the following error:
>>
>> shcond
>> terminate called after throwing an instance of 'NEWMAT::IndexException'
>> /usr/local/fsl/bin/run_first: line 161: 14511 Aborted
>> ${FSLDIR}/bin/first
>> -i $image -l $trmat -m $model -k $outname -n $modes $verbose $comExtras
>>
>> I've tried recreating my flirt-registered .mat and original run_first
>> .bvars
>> files with the new software and then using these in the command, but I
>> receive the same error.
>>
>> Is this a common problem due to the switch or something that might be
>> wrong
>> with my data? NOTE - I am trying to do a R_Thal_given_R_Late segmentation,
>> and I had no problem at all running the L_Thal_given_L_Late segmentation
>> with the older version.
>>
>> Thanks,
>>
>> Zac
>>
>> ------------------
>>
>> [log in to unmask]
>> Brain Imaging Lab
>> Psychiatry Associates
>> Dartmouth-Hitchcock Medical Center
>>
>>
>>