Thank you, Tim.

I let it run longer and it completed this time and I got the desired output. I must have been impatient and cancelled it too early before!

I doubled the speed by running it on my two processors. I think I'm going to try to parallelize to other computers as well.

I have still only successfully run it on the subj1 sample data. For my data, it does not create datastart and dataend, retstart and retend for my first and last slices. I will try running it on some of my other data and see what happens.

Thanks again,
Marc


On 6/9/07, Tim Behrens <[log in to unmask]> wrote:
Marc, 
This is exactly what a single slice output should look like. 
When all the slices are finished they are merged together into  files called merged_...

 
You are looking for a directory that doesn't look like this!

 
T

On 9 Jun 2007, at 00:21, Marc Dubin wrote:

Hi Tim,

 
In follow up to my previous email, I also wanted to let you know that the bedpost output is very different than the output for the sample data (subj1.bedpost).

 
For example, for the data_slice_0010 directory, I get:

 
   975111 Jun  6 01:45 S0_samples.nii.gz
   971621 Jun  6 01:45 d_samples.nii.gz
         183 Jun  6 01:45 dataend
         183 Jun  6 01:45 datastart
1010243 Jun  6 01:45 f_samples.nii.gz
      627 Jun  8 19:16 fdt.log
       429 Jun  6 01:36 logfile
    995138 Jun  6 01:45 ph_samples.nii.gz
       318 Jun  6 01:45 retend
       318 Jun  6 01:45 retstart
    991613 Jun  6 01:45 th_samples.nii.gz

I don't get any mean_*, merged_* or nodif_* files.

 
Best,
Marc


 

 

 
nifti_image_read(/Users/marcDocuments/Dir.bedpost/
diff_slices/data_slice_0010/
On 6/7/07, Tim Behrens < [log in to unmask]> wrote:
Hi - the syntax error always occurs (see previous postings here). The debugging-style logging also always occurs - it is fine, except for taking up disk space etc. 
Both of these "features" are cleaned up for the next release.

 
T

 

On 7 Jun 2007, at 14:43, Marc Dubin wrote:

Tim,

Thank you for your help. I will start by trying to find the slice that's causing problems. However, I think I may be doing something wrong in bedpost in general because when I try to run it on the sample data in fdt_subj1, the program ends with a syntax error (293). I have just made a more detailed posting about his. I understand that you are busy and will patiently await your reply whenever you have time. In the meantime, if I solve my own problem I will be sure to let the group know.

Thanks again,
Marc

On 6/7/07, Tim Behrens < [log in to unmask]> wrote:
It sounds like a single slice has crashed and not finished.

Could you look in the slice directories to see which one has crashed
(the outputs will not be in the directory), and see if there is
anything different about the data from that slice. if not, can you
send us that slice (data_slice_blah and nodif_brain_mask_slice...)
and the bvecs and bvals by uploading them to
http://www.fmrib.ox.ac.uk/cgi-bin/upload.cgi
and telling us the number.

I will take a look as soon as I can, but I'm afraid I am flying to
HBM very soon, so I cannot guarantee to do it before getting back.

T



On 6 Jun 2007, at 23:40, Marc Dubin wrote:

> Hello,
>
> When I run bedpost on my macbook pro intel core 2 duo 2.16 Ghz,
> on my 30 slice data with 26 diffusion directions, it appears to finish
> in about 2 hours (appears that all outpt is there in the directory
> structure)
> and then enter an infinite loop, printing to screen "29 slices
> process" about
> every half hour. Does anyone know if it's done or still doing
> something important?
>
> Best,
> Marc



--
646 831 8886
[log in to unmask]




--
646 831 8886
[log in to unmask]




--
646 831 8886
[log in to unmask]