Re: [HCP-Users] visualise and manipulate tractography seeded from voxels

[Timothy Coalson]

wb_command -cifti-create-dense-scalar can make a file for this purpose,
take the same ROI as you gave the tractography, and import it as a label
volume with a name of "CEREBRAL_WHITE_MATTER_LEFT" or _RIGHT.  I forget the
details of how voxel seeds get specified to the tractography, you may have
to be careful about how they are ordered when you run the tractography
(there is no simple way to make a cifti file with a different voxel order
of the same ROI).

The "template" part is just how the file gets used (like "roi" in "roi
volume"), and use that as the structure label volume.  The "template" part
is not a special format, it is just saying how the file gets used, it is
like "ROI" in "ROI volume" - same format, different purpose.  In this case,
all it needs is to know which voxel each index in the tractography output
corresponds to, and it is written to take that information from a cifti
mapping.

Tim


On Thu, May 31, 2018 at 4:51 AM, Colin Reveley  wrote:

> Hi - a while back I had a lot of success using workbench in a study of
> tractography and anatomy. I seeded tractography from the midthickness
> cortical surface of mostmortem macaque data. Then I used
>
> wb_command-convert-matrix4-to-workbench-sparse
>
> and
>
> wb_command -convert-matrix4-to-matrix2
>
> to get a matrix of gray matter vertices to white matter voxels, and
> also to get trajTEMP.wbsparse files from gray matter ROI to display
> visually in workbench.
>
> This is straigtforward using surfaces as seeds. Now I want to
> expand this analysis to entities within the white matter, i.e. fiber
> budles.
>
> I want to seed from a set of voxels, and then make a
> trajTEMP.wbsparse. That would be the start, in general I want to
> implement, expand and enrich the ideas here
> https://elifesciences.org/articles/35237.
>
> convert-matrix4-to-workbench-sparse will take voxels as seeds. What I
> need is a matrix of seed voxels by brain voxels instead of vertices by
> voxels.
>
> But to make this I need a cifti template. There is no way to generate
> one using wb_command that I can see.
>
> I think I need to make this using the matlab code on github, and the
> cifti specification. But I am finding it very challenging to do this.
>
> Is there any kind of tutorial, or gentle documentation on this?
>
> best wishes
>
> Colin
>
> Colin Reveley
> Neuroecology and Anatomy lab
> Wellcome Centre for Integrative Neuroimaging
> Centre for Functional MRI of the Brain (FMRIB)
> Nuffield Department of Clinical Neurosciences
> John Radcliffe Hospital
> University of Oxford
> John Radcliffe Hospital
> University of Oxford
>
> ___
> HCP-Users mailing list
> HCP-Users@humanconnectome.org
> http://lists.humanconnectome.org/mailman/listinfo/hcp-users
>

___
HCP-Users mailing list
HCP-Users@humanconnectome.org
http://lists.humanconnectome.org/mailman/listinfo/hcp-users

Re: [HCP-Users] error running hcp_fix

[Marta Moreno]

Thanks. I can install randomForest_4.6-14 instead. IS there any specific reason 
for needing version 4.6-12 to run hip_fix?



***
Leah Moreno, PhD
Associate Research Scientist
Division of Experimental Therapeutics
Department of Psychiatry
Columbia University Medical Center
1051 Riverside Drive, Unit 21
New York, NY 10032
phone: (914) 218-7311
email: mm4...@cumc.columbia.edu
email: mmor...@nyspi.columbia.edu

> On May 30, 2018, at 7:52 PM, Timothy Coalson  wrote:
> 
> That is the same line I used to install randomForest in a docker container.  
> I would guess that those look like assembly directives.  Based on that, I 
> would guess that either randomForest uses some assembly code, and it doesn't 
> support the OS X compiler you are using, or that you somehow have a broken 
> compiler setup (less likely).
> 
> The main possibilities I see:
> 
> 1) do your processing using linux instead of OS X (VM, container, or a native 
> install)
> 2) try installing gcc and gfortran on OS X, and getting R to use them to 
> build packages (probably messy, may need to rebuild R itself, we haven't 
> tested it)
> 
> Tim
> 
> 
> On Wed, May 30, 2018 at 8:59 AM, Marta Moreno  > wrote:
> Dear Tim,
> 
>  How did you install randomForest' version 4.6.12? I am getting the following 
> error (see below). Do I need R-devel as well to install randomForest? Please 
> advice this should be the last step before having hcp_fix running properly. 
> 
> install_version("randomForest", version = "4.6-12", repos = 
> "http://cran.us.r-project.org ”)
> 
> (…)
> 
> /var/tmp//ccP5LbYW.s:2671:2: error: unknown directive
> .stabn  224,0,0,LBE141
> ^
> /var/tmp//ccP5LbYW.s:2673:9: error: unsupported directive '.stabs'
> .stabs  "",36,0,0,Lscope8-LFBB8
> ^
> /var/tmp//ccP5LbYW.s:2697:9: error: unsupported directive '.stabs'
> .stabs  "",100,0,0,Letext0
> ^
> /var/tmp//ccP5LbYW.s:2835:11: warning: section "__textcoal_nt" is deprecated
> .section __TEXT,__textcoal_nt,coalesced,pure_instructions
>  ^  ~
> /var/tmp//ccP5LbYW.s:2835:11: note: change section name to "__text"
> .section __TEXT,__textcoal_nt,coalesced,pure_instructions
>  ^  ~
> make: *** [rfsub.o] Error 1
> ERROR: compilation failed for package ‘randomForest’
> * removing 
> ‘/Library/Frameworks/R.framework/Versions/3.5/Resources/library/randomForest’
> Installation failed: Command failed (1)
> 
> Thanks a lot!
> 
> -L
> 
> ***
> Leah Moreno, PhD
> Associate Research Scientist
> Division of Experimental Therapeutics
> Department of Psychiatry
> Columbia University Medical Center
> 1051 Riverside Drive, Unit 21 
> 
> New York, NY 10032 
> 
> phone: (914) 218-7311
> email: mm4...@cumc.columbia.edu 
> email: mmor...@nyspi.columbia.edu 
> 
>> On May 27, 2018, at 11:20 PM, Marta Moreno > > wrote:
>> 
>> Thanks, Tim! I followed your advice and I am almost there… How did you 
>> install randomForest' version 4.6.12? I am getting the following error: 
>> 
>> install_version("randomForest", version = "4.6-12", repos = 
>> "http://cran.us.r-project.org ”)
>> 
>> (…)
>> 
>> /var/tmp//ccP5LbYW.s:2671:2: error: unknown directive
>> .stabn  224,0,0,LBE141
>> ^
>> /var/tmp//ccP5LbYW.s:2673:9: error: unsupported directive '.stabs'
>> .stabs  "",36,0,0,Lscope8-LFBB8
>> ^
>> /var/tmp//ccP5LbYW.s:2697:9: error: unsupported directive '.stabs'
>> .stabs  "",100,0,0,Letext0
>> ^
>> /var/tmp//ccP5LbYW.s:2835:11: warning: section "__textcoal_nt" is deprecated
>> .section __TEXT,__textcoal_nt,coalesced,pure_instructions
>>  ^  ~
>> /var/tmp//ccP5LbYW.s:2835:11: note: change section name to "__text"
>> .section __TEXT,__textcoal_nt,coalesced,pure_instructions
>>  ^  ~
>> make: *** [rfsub.o] Error 1
>> ERROR: compilation failed for package ‘randomForest’
>> * removing 
>> ‘/Library/Frameworks/R.framework/Versions/3.5/Resources/library/randomForest’
>> Installation failed: Command failed (1)
>> 
>> Thanks!
>> 
>> -L
>> 
>>> On May 27, 2018, at 9:12 PM, Timothy Coalson >> > wrote:
>>> 
>>> The way I solved this was to use the very latest R version (cran mirrors 
>>> contain debian and ubuntu packages, see 
>>> http://cran.wustl.edu/bin/linux/ubuntu/ 
>>>  , and they also have OS X 
>>> binaries: http://cran.wustl.edu/bin/macosx/