Hi Again,
(apologies for the many posts): I should add that none of this happens
(and everything works fine - no warning and small clusters which look
reasonable) if I use version 5.3.
Thank you,
Nicola
On 7/22/2016 2:10 PM, Nicola Toschi wrote:
> Hi,
>
> thanks - I tried using "abs" as well as a higher threshold, however the
> large, whole-brain cluster (as well as the warning messages) still remains.
>
> Do the warnings:
>
> WARNING: unrecognized mri_glmfit cmd option mri_glmfit.bin
> WARNING: 251446 NaNs found in volume
>
> give any clues?
>
> The strange thing is that everything is fine (small clusters and no
> warning) when I use t-tests )i.e. just the first or second line of the F
> test).
>
> Thanks a lot for your time,
>
> Nicola
>
>
> On 07/21/2016 11:53 PM, Douglas N Greve wrote:
>> Try using abs instead of pos. abs (absolute) is more appropriate since
>> an F test is unsigned. Also, I would go with a threshold of 2 and not
>> 1.3. I'm just now finding out that such a low threshold can cause a
>> lot of false postives on real data.
>>
>>
>> On 07/21/2016 05:32 PM, Nicola Toschi wrote:
>>> Hi,
>>>
>>> thanks for your reply. The mri_glmfit-sim was pasted below, here it
>>> is the dereferenced version:
>>>
>>>
>>> /usr/local/freesurfer6dev/bin/mri_glmfit-sim --glmdir
>>> design_3groups_AgeGenderIq_area_lh_fwhm0_pial --cache 1.3 pos
>>>
>>>
>>> thanks again,
>>>
>>> Nicola
>>>
>>> On 7/21/2016 11:13 PM, Douglas N Greve wrote:
>>>> That is the mri_glmfit command, I need the mri_glmfit-sim. Also,
>>>> please do not include variables, just the full command with all
>>>> dereferenced args
>>>>
>>>> On 07/21/2016 03:28 PM, Nicola Toschi wrote:
>>>>> Hi,
>>>>>
>>>>> here it is (I also pasted the mri_glmfit line):
>>>>>
>>>>> /odir=${prefix}_${meas}_${h}_fwhm${s}_${surf} #name output directory//
>>>>> //
>>>>> //mri_glmfit --y ${h}.fwhm${s}.${list}.${meas}.mgh --X ${matrix} \//
>>>>> //--C ${C1} \//
>>>>> //--glmdir ${odir} --surf fsaverage ${h} ${surf}//
>>>>> //
>>>>> //*/usr/local/freesurfer6dev/bin/mri_glmfit-sim --glmdir ${odir}
>>>>> --cache ${t} pos*//*
>>>>> */
>>>>> Note: it is the same code i use for the t-tests, which appear to
>>>>> work fine.
>>>>>
>>>>> The content of the file ${C1} is
>>>>>
>>>>> 1 -1 0 0 0 0
>>>>> 1 0 -1 0 0 0
>>>>>
>>>>>
>>>>> Thanks in advance!
>>>>>
>>>>> Nicola
>>>>>
>>>>>
>>>>>
>>>>> On 7/21/2016 5:38 PM, Douglas N Greve wrote:
>>>>>> what is your mri_glmfit-sim command line?
>>>>>>
>>>>>> On 07/20/2016 05:56 PM, Nicola Toschi wrote:
>>>>>>> Hi List,
>>>>>>>
>>>>>>> I am getting a couple of strange error when running a 3-group
>>>>>>> F-test.
>>>>>>>
>>>>>>> WARNING: unrecognized mri_glmfit cmd option mri_glmfit.bin
>>>>>>> WARNING: 251446 NaNs found in volume
>>>>>>> analysis/Ftest/cache.th20.pos.sig.cluster.mgh...
>>>>>>>
>>>>>>> And as a result, I consistently get huge clusters:
>>>>>>>
>>>>>>> # ClusterNo Max VtxMax Size(mm^2) MNIX MNIY MNIZ CWP
>>>>>>> CWPLow CWPHi NVtxs WghtVtx Annot
>>>>>>> 1 inf 0 63247.45 -38.8 -19.0 66.9 0.00010
>>>>>>> 0.00000 0.00020 98722 inf precentral
>>>>>>>
>>>>>>> However, this doesn't happen when running t-tests on the same data.
>>>>>>> Still, I think my F-contrast is correct (see below).
>>>>>>>
>>>>>>> Thanks in advance for any advice!
>>>>>>>
>>>>>>> Nicola
>>>>>>>
>>>>>>> PS:
>>>>>>>
>>>>>>> my F-contrast looks like this:
>>>>>>>
>>>>>>> 1 -1 0 0 0 0
>>>>>>> 1 0 -1 0 0 0
>>>>>>>
>>>>>>> and my design matrix looks like this:
>>>>>>>
>>>>>>> 0 0 1 1 179 110
>>>>>>> 0 0 1 1 193 103
>>>>>>> 0 0 1 1 176 108
>>>>>>> 0 0 1 1 198 94
>>>>>>> 0 1 0 1 186 87
>>>>>>> 0 1 0 1 217 83
>>>>>>> .....
>>>>>>>
>>>>>>> and my versions are
>>>>>>>
>>>>>>> # $Id: mri_surfcluster.c,v 1.57 2014/03/06 17:02:46 greve Exp $
>>>>>>> # $Id: mrisurf.c,v 1.776 2015/12/17 18:09:34 fischl Exp $
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> 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
