Email discussion lists for the UK Education and Research communities |
|
 |
Archive Search
|
|
WWW.JISCMAIL.AC.UK ( 40 Matches )
Subject | From | Date | Time |
| So, what I did to test everything out was taking an atlas, adding a random number with mean 0 and sigma 1 to the whole area and saving 100 such random "scans". Then I generated a scan adding 5 to this atlas. I first performed a two groups t-test with equal variance in spm5 and spm8. Both versions give a design matrix which is [1 0;0 1] and the spmT image is an uniform "5". Perfectly, as expected! If I select "unequal" instead I get an undisplayable design Matrix ("image CData cannot be complex") and I cannot even define contrasts to compute the con or spmT files in SPM8. In SPM5 I get a uniform spmT with a value of 16.19, and the design Matrix is [0.99 0;0 3.30]. I've briefly looked at the database of past scans and it initially appears that all of ours design matrixes have these values. |
| So, what I did to test everything out was taking an atlas, adding a random number with mean 0 and sigma 1 to the whole area and saving 100 such random "scans". Then I generated a scan adding 5 to this atlas. I first performed a two groups t-test with equal variance in spm5 and spm8. Both versions give a design matrix which is [1 0;0 1] and the spmT image is an uniform "5". Perfectly, as expected! If I select "unequal" instead I get an undisplayable design Matrix ("image CData cannot be complex") and I cannot even define contrasts to compute the con or spmT files in SPM8. In SPM5 I get a uniform spmT with a value of 16.19, and the design Matrix is [0.99 0;0 3.30]. I've briefly looked at the database of past scans and it initially appears that all of ours design matrixes have these values. |
From: Guillaume Flandin <[log in to unmask]>
Subject: Re: Image CData can not be complex
Comments: To: Cristina Scarpazza <[log in to unmask]>
> Error using image
> Image CData can not be complex
> |
From: Cristina Scarpazza <[log in to unmask]>
Subject: Image CData can not be complex
MIME-Version: 1.0
Error using image
Image CData can not be complex |
Error using ==> image
Image CData can not be complex |
Error using ==> image
Image CData can not be complex |
Error using ==> image
Image CData can not be complex. |
Subject: Re: How does this error come in paired t-test in SPM5? ----"Image
CData can not be complex"
Mime-Version: 1.0
抄送： SPM
主题： Re: [SPM] How does this error come in paired t-test in SPM5? ----"Image CData can not be complex"
Hello,
this has been discussed a few times on the SPM mailing list:
https://www.jiscmail.ac.uk/cgi-bin/wa.exe?S2=SPM&q=CData+complex
ie, for a paired t-test, you should set "Equal" Variance.
> Error using ==> image
> Image CData can not be complex
> Error in ==> spm_conman at 975 |
Subject: Re: How does this error come in paired t-test in SPM5? ----
"Image CData can not be complex"
Comments: To: wlys111 <[log in to unmask]>
this has been discussed a few times on the SPM mailing list:
https://www.jiscmail.ac.uk/cgi-bin/wa.exe?S2=SPM&q=CData+complex
ie, for a paired t-test, you should set "Equal" Variance.
> Error using ==> image
> Image CData can not be complex
> Error in ==> spm_conman at 975 |
Subject: How does this error come in paired t-test in SPM5? ---- "Image
CData can not be complex"
Mime-Version: 1.0
Error using ==> image
Image CData can not be complex
Error in ==> spm_conman at 975 |
>> Error using ==> image
>> Image CData can not be complex
> [...]
> another format, this would explain that the error occurs after
> smoothing. CData is involved in plotting and defining color properties,
> so why it occurs, I have no idea. |
> Error using ==> image
> Image CData can not be complex
[...]
another format, this would explain that the error occurs after
smoothing. CData is involved in plotting and defining color properties,
so why it occurs, I have no idea. |
Error using ==> image
Image CData can not be complex |
> Error using ==> image
> Image CData can not be complex.
> |
Error using ==> image
Image CData can not be complex. |
> Error using ==> image
> Image CData can not be complex
> |
Error using ==> image
Image CData can not be complex |
I had luckily avoid this problem by restarting Matlab, though I tried to clean up the memory and path.
I did a brief tracking of this problem, and found the SPM.xCon.eidf was calculated to be a complex number. But I didn't track why and how this could happen.
Error using ==> image
Image CData can not be complex. |
Error using ==> image
Image CData can not be complex. |
Error using ==> image
Image CData can not be complex |
Organization: Freiburg Brain Imaging
Subject: Re: Error when running paired t-test on smoothed data, complex nKX
In-Reply-To: <[log in to unmask]>
> estimation. Also, I checked, and my SPM.xX.nKX does in fact have
> complex values!
> Few more pieces of perhaps useful info: error occurs when I run
> Error using ==> image
> Image CData can not be complex.
> |
From: layla sara <[log in to unmask]>
Subject: Error when running paired t-test on smoothed data, complex nKX
MIME-Version: 1.0
pasted in the bottom of this message. No errors appear during estimation.
Also, I checked, and my SPM.xX.nKX does in fact have complex values!
Few more pieces of perhaps useful info: error occurs when I run analysis
Error using ==> image
Image CData can not be complex. |
From: layla sara <[log in to unmask]>
Subject: Error when running paired t-test on smoothed data, complex nKX
MIME-Version: 1.0
pasted in the bottom of this message. No errors appear during estimation.
Also, I checked, and my SPM.xX.nKX does in fact have complex values!
Few more pieces of perhaps useful info: error occurs when I run analysis
Error using ==> image
Image CData can not be complex. |
> it seems you end up with some complex hyperparameter estimate... This is
> probably caught in the new SPM version right after spm_reml.
> But why they
> get complex is another question. It seems that ReML is converging very
> slowly... Try increasing the number of covariance components slowly, e.g.
Yup that's exactly what I did. Unqeual variance for time only :-( So we either
stick without hyperparameters or complex design matrices (Unfortunately I
have a few difficulties interprating the latter ones ;)
> > The version before that estimates the model fine but I cannot get results
> > because SPM.xX.nKX becomes complex (which I think is strange).
> >
> > Error using ==> image
> > Image CData can not be complex..
> >
> > Error using ==> image
> > Image CData can not be complex..
> > |
it seems you end up with some complex hyperparameter estimate... This is
probably caught in the new SPM version right after spm_reml. But why they
get complex is another question. It seems that ReML is converging very
slowly... Try increasing the number of covariance components slowly, e.g.
> The version before that estimates the model fine but I cannot get results
> because SPM.xX.nKX becomes complex (which I think is strange).
>
> Error using ==> image
> Image CData can not be complex..
>
> Error using ==> image
> Image CData can not be complex..
> |
The version before that estimates the model fine but I cannot get results
because SPM.xX.nKX becomes complex (which I think is strange).
Error using ==> image
Image CData can not be complex..
Error using ==> image
Image CData can not be complex.. |
From: Joern Diedrichsen <[log in to unmask]>
Subject: Re: RobustWLS and complex data error message
Comments: To: marco tettamanti <[log in to unmask]>
The complex data error message arises from negative variances
estimates from the RobustWLS method. Unfortunately, it is not a
>
> xX.xKXs.X, xX.X, SPM.xX.K, and SPM.xX.W are all complex.
>
>
> SPM.xVi.V in turn is not complex.
>
> Also pointing to your suspect of a bug in the rWLS toolbox, I have
> now found out that the "complex data" error only occurs when I
> choose 'WLS' as as (co-) variance structure for serial correlations
>> I can think of a couple of things to try:
>> Check whether xX.xKXs.X has complex elements, or whether the
>> display scaling is just messing things up -- I expect the former,
>> could try loading the SPM.mat after estimation, removing the
>> complex values in xX.nKX and then resaving; the results might then
>> work alright, although you'd have to ignore the displayed matrix.
>> If xX.xKXs.X is complex, double check that the original xX.X is
>> what you expect, and then investigate SPM.xX.K and SPM.xX.W, and
>> if the latter is complex, then also look at SPM.xVi.V. If that is
>> where the problem appears, then it sounds like ReML has gone awry,
>>> contrast effects clicking on the "Results" button, I get an
>>> "Image CDdata can not be complex" error (see error message
>>> attached below). Oddly, this only happens for a subset of subjects.
>>>
>>> When I do this, I find that the nKX structure is indeed complex
>>> for the subjects in which I encounter the problem and that it is
>>> not complex for the "good" subjects.
>>>
>>> always run fine, whereas the "bad" subjects always give the
>>> "complex data" error. Also, rWLS convergence does not seem to be
>>> a problem.
>>> However, an "equivalent" statistical model estimation in SPM5
>>> (i.e. outside the RWLS toolbox) does not produce any "complex
>>> data" error, even in the bad subjects.
>>> Error using ==> image
>>> Image CData can not be complex..
>>>
>>> Error using ==> image
>>> Image CData can not be complex..
>>> |
From: marco tettamanti <[log in to unmask]>
Subject: Re: RobustWLS and complex data error message
Comments: To: Ged Ridgway <[log in to unmask]>
xX.xKXs.X, xX.X, SPM.xX.K, and SPM.xX.W are all complex.
SPM.xVi.V in turn is not complex.
Also pointing to your suspect of a bug in the rWLS toolbox, I have now
found out that the "complex data" error only occurs when I choose 'WLS'
as as (co-) variance structure for serial correlations and NOT when I
>
> Check whether xX.xKXs.X has complex elements, or whether the display
> scaling is just messing things up -- I expect the former, unfortunately,
> but if it turns out it is just the latter, then you could try loading
> the SPM.mat after estimation, removing the complex values in xX.nKX and
> then resaving; the results might then work alright, although you'd have
>
> If xX.xKXs.X is complex, double check that the original xX.X is what you
> expect, and then investigate SPM.xX.K and SPM.xX.W, and if the latter is
> complex, then also look at SPM.xVi.V. If that is where the problem
> appears, then it sounds like ReML has gone awry, even if the estimation
>> MakeCon), but when I attempt to display the contrast effects clicking
>> on the "Results" button, I get an "Image CDdata can not be complex"
>> error (see error message attached below). Oddly, this only happens for
>>
>> When I do this, I find that the nKX structure is indeed complex for
>> the subjects in which I encounter the problem and that it is not
>> complex for the "good" subjects.
>>
>> enter the unsmoothed data or not: the "good" subjects always run fine,
>> whereas the "bad" subjects always give the "complex data" error. Also,
>> rWLS convergence does not seem to be a problem.
>> However, an "equivalent" statistical model estimation in SPM5 (i.e.
>> outside the RWLS toolbox) does not produce any "complex data" error,
>> even in the bad subjects.
>> Error using ==> image
>> Image CData can not be complex..
>>
>> Error using ==> image
>> Image CData can not be complex..
>> |
From: Ged Ridgway <[log in to unmask]>
Subject: Re: RobustWLS and complex data error message
Comments: To: marco tettamanti <[log in to unmask]>
Check whether xX.xKXs.X has complex elements, or whether the display
scaling is just messing things up -- I expect the former,
unfortunately, but if it turns out it is just the latter, then you
could try loading the SPM.mat after estimation, removing the complex
values in xX.nKX and then resaving; the results might then work
If xX.xKXs.X is complex, double check that the original xX.X is what
you expect, and then investigate SPM.xX.K and SPM.xX.W, and if the
latter is complex, then also look at SPM.xVi.V. If that is where the
problem appears, then it sounds like ReML has gone awry, even if the
> MakeCon), but when I attempt to display the contrast effects clicking on
> the "Results" button, I get an "Image CDdata can not be complex" error
> (see error message attached below). Oddly, this only happens for a
>
> When I do this, I find that the nKX structure is indeed complex for the
> subjects in which I encounter the problem and that it is not complex for
> the "good" subjects.
> enter the unsmoothed data or not: the "good" subjects always run fine,
> whereas the "bad" subjects always give the "complex data" error. Also,
> rWLS convergence does not seem to be a problem.
> However, an "equivalent" statistical model estimation in SPM5 (i.e.
> outside the RWLS toolbox) does not produce any "complex data" error,
> even in the bad subjects.
> Error using ==> image
> Image CData can not be complex..
>
> Error using ==> image
> Image CData can not be complex..
> |
From: marco tettamanti <[log in to unmask]>
Subject: RobustWLS and complex data error message
In-Reply-To: <[log in to unmask]>
MakeCon), but when I attempt to display the contrast effects clicking on
the "Results" button, I get an "Image CDdata can not be complex" error
(see error message attached below). Oddly, this only happens for a
When I do this, I find that the nKX structure is indeed complex for the
subjects in which I encounter the problem and that it is not complex for
the "good" subjects.
enter the unsmoothed data or not: the "good" subjects always run fine,
whereas the "bad" subjects always give the "complex data" error. Also,
rWLS convergence does not seem to be a problem.
However, an "equivalent" statistical model estimation in SPM5 (i.e.
outside the RWLS toolbox) does not produce any "complex data" error,
even in the bad subjects.
Error using ==> image
Image CData can not be complex..
Error using ==> image
Image CData can not be complex.. |
> Image CData can not be complex.
> Error in ==> spm_conman at 960 |
Error using ==> image
Image CData can not be complex. |
> Error using ==> image
> Image CData can not be complex..
>
> Error using ==> image
> Image CData can not be complex..
> |
Error using ==> image
Image CData can not be complex..
Error using ==> image
Image CData can not be complex.. |
??? Error using ==> image
Image CData can not be complex.
??? Error using ==> image
Image CData can not be complex. |
design matrix is empty and an error appear in the matlab window " error
using ==> image image CData cannot be a complex error in ...
spm_DesRep.m line 622" |
Error using ==> image
Image CData can not be complex. |
??? Error using ==> image
Image Cdata can not be complex.
??? Error using ==> image
Image Cdata can not be complex. Etc etc ...
We presume there was negative covariance producing the concerned complex
numbers?
Can the non-sphericity scripts be modified to deal with the presence of
complex numbers on the rare occasions they are generated? |
The problem reported is that matlab tried to show an image containing
complex values. It appears that your design matrix contains complex
values. No idea how they got there. Did you make the design matrix in
> ??? Error using ==> image
> Image CData can not be complex.
> |
??? Error using ==> image
Image CData can not be complex. |
|
|
|
