30 Jul
2012
30 Jul
'12
5:34 p.m.
Hi Sergio - Sorry, if this occurs independently of trac-all, whenever you run bedpostx directly, you're better off asking for help on the FSL list. We've never run into this problem with bedpostx up till now.
a.y
On Mon, 30 Jul 2012, Sergio Medina wrote:
Nope, I have the same files as before :(
2012/7/27 Anastasia Yendiki ayendiki@nmr.mgh.harvard.edu
The monitoring script that bedpostx runs and tells you if things have or haven't finished properly can't always be trusted. So if that's the only error message you see, I wouldn't worry about it. Have the *ph1* and *th1* files appeared this time? On Fri, 27 Jul 2012, Sergio Medina wrote: What a coincide! Our teams have both chosen the same release cycle strategy :P I ran bedpostx_postproc and I've got so many errors that makes me think the problem is before this step (see output in attached n1-bedpostx-postproc.txt). So, I ran again bedpostx on the dmri directory, which the first run *seemed* to finish and didn't throw any error (I attached the output in the first email). This run did finish, and it did throw errors (full output in n2-bedpostx-dmri.txt): [...] slice 63 has already been processed wc: /volatile/tracula/results/Diff001/dmri.bedpostX/commands.txt: No such file or directory Queuing post processing stage 64 slices processed For some reason the bedpostX process DOES NOT appear to have successfully completed. Please examine your results carefully. kill: 276: No such process Then I started wondering whether I do have the right data to actually run bedpostx, so I run the script that checks that: bedpostx_datacheck, the output (n3-bedpostx-datacheck.txt) doesn't show any signs of errors so I would like to assume everything is ok with the input data... Any idea? Thanks, once again. Sergio. 2012/7/27 Anastasia Yendiki <ayendiki@nmr.mgh.harvard.edu> Hi Sergio - As with any imminent new version release, it's happening "soon". I'd say in a month, I'm just not sure a month from when :P The fact that the temporary files (everything with *slice* in the name) are still in your bedpostx directory means that bedpostx didn't finish running properly. You can try running the last step of bedpostx and see if you get any error messages that give you any clues: bedpostx_postproc.sh /volatile/tracula/results/Diff001/dmri Normally this step merges the individual slices into volumes and deletes the temporary files. If all else fails, I'd ask the FSL gurus. Hope this helps, a.y On Fri, 27 Jul 2012, Sergio Medina wrote: Anastasia, Thanks for such a quick reply! Please find my answers inline... 2012/7/26 Anastasia Yendiki <ayendiki@nmr.mgh.harvard.edu> Hi Sergio - Thank you for your kind words, and also thank you for taking the time to search the archives for answers! The first problems you had with running bedpostx through trac-all have to do with trac-all assuming an older version of FSL than the current one, and this'll be fixed in the upcoming release of trac-all. Indeed the thing to do in the meantime is to run bedpostx by itself, as you did. I apologize for the inconvenience. Great! By any chance, do you know when that release will happen? About the problem you had when you ran bedpostx directly: do you get the other outputs of bedpostx, listed here? Are there any other, temporary files left over as well? http://www.fmrib.ox.ac.uk/fsl/fdt/fdt_bedpostx.html I have all of them, except for nodif_brain. And I only have the ones with i=2 while I should have the ones with i=1 too. Attached you can find two files, the output of ls -Rah on both dmri and dmri.bedpostX directories. The merged_ph?samples are the orientation angles of the each of the two anisotropic compartments ("sticks", "fibers", whatever you want to call them) that bedpostx fit in each voxel. So "1" is the orientation angle of the largest stick/fiber, and "2" the one of the 2nd largest. If there's a "2", there should be a "1"! But there are not :( Also, I looked at your dmrirc file and noticed that you decreased the # of control points b/c you hate waiting :) I just wanted to warn you that decreasing that # might make you wait longer b/c make fitting a spline with fewer control points to initialize the algorithm is sometimes more challenging. Hehe, thanks! I changed it back to 5 now... Thanks! Sergio. Hope this helps, a.y On Thu, 26 Jul 2012, Sergio Medina wrote: Hi! First of all let me thank you for having put at our disposal such great tools as FreeSurfer and Tracula! I am trying to run Tracula on the sample data that is on the website, but I've been running into some issues. I am attaching the configuration file just in case (dmrirc_single_subject-serch). The first step "trac-all -prep" seems to run fine. The first problem comes with the second step "trac-all -bedp". The complete trace is included in the attached file 1-error-trac-all-bedp.txt After finding this on theFreeSurfermailinglisthttps://mail.nmr.mgh.harvard.edu/pipermail//freesurfer /2012-May/ 024042. html I changed the lines ${BPDIR}/fsl_sub_seychelles for ${FSL_BIN}/fsl_sub in the bedpostx_seychelles and I could continue, until the script seems to be stuck Queuing parallel processing stage /volatile/tracula/results/Diff001/dmri.bedpostX/monitor 0 slices processed 1 slices processed 1 slices processed 1 slices processed <== this line keeps on repeating Following a suggestion I found on this other archived email http://www.mail-archive.com/freesurfer@nmr.mgh.harvard.edu/msg20815.html I ran bedpostx on the dmri directory and seemed to succeed. I am attaching the output (2-bedpostx-dmri.txt). Then I tried to run trac-all -path but it stops with the following error: niiRead(): error opening file /volatile/tracula/results/Diff001/dmri.bedpostX/merged_ph1samples.nii.gz ERROR: Could not read /volatile/tracula/results/Diff001/dmri.bedpostX/merged_ph1samples.nii.gz I searched for that file everywhere but it doesn't exist. The closest I found was results/Diff001/dmri.bedpostX/merged_ph2samples.nii.gz Just in case I tried once again to run trac-all -bedp and this time, after having run bedpostx on the dmri directory, I get the following error (the whole output is in the attached 4-error-trac-all-bedp-2.txt file): [...] slice 62 has already been processed slice 63 has already been processed wc: /volatile/tracula/results/Diff001/dmri.bedpostX/commands.txt: No such file or directory Queuing post processing stage 64 slices processed For some reason the bedpostX process DOES NOT appear to have successfully completed. Please examine your results carefully. kill: 390: No such process I tried searching for a file named commands.txt but I couldn't find any. Here is the output of bugr: FREESURFER_HOME: /i2bm/local/freesurfer Build stamp: freesurfer-Linux-centos4_x86_64-stable-pub-v5.1.0 Debian version: squeeze/sid Kernel info: Linux 2.6.38-14-generic x86_64 and the output of source $FREESURFER_HOME/SetUpFreeSurfer.sh -------- freesurfer-Linux-centos4_x86_64-stable-pub-v5.1.0 -------- Setting up environment for FreeSurfer/FS-FAST (and FSL) FREESURFER_HOME /i2bm/local/freesurfer FSFAST_HOME /i2bm/local/freesurfer/fsfast FSF_OUTPUT_FORMAT nii.gz SUBJECTS_DIR /volatile/tracula/data/buckner_data/tutorial_subjs/ MNI_DIR /i2bm/local/freesurfer/mni FSL_DIR /usr/share/fsl/4.1 Thanks in advance for your help, any hint would be highly appreciated. Best, Sergio. -- Sergio Medina Engineer INRIA Parietal - Neurospin Bât 145, Point Courrier 156 91191 Gif/Yvette, France Tel.: (+33) 1 69 08 80 85 _______________________________________________ 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. -- Sergio Medina Engineer INRIA Parietal - Neurospin Bât 145, Point Courrier 156 91191 Gif/Yvette, France Tel.: (+33) 1 69 08 80 85 -- Sergio Medina Engineer INRIA Parietal - Neurospin Bât 145, Point Courrier 156 91191 Gif/Yvette, France Tel.: (+33) 1 69 08 80 85-- Sergio Medina Engineer INRIA Parietal - Neurospin Bât 145, Point Courrier 156 91191 Gif/Yvette, France Tel.: (+33) 1 69 08 80 85