It might be an issue with the degrees of freedom. How many subjects do
you have and what type of design have you specified?
On 07/06/18 13:05, Hellen Mathei Della-Justina wrote:
> Dear SPM experts,
> I'm running a 2nd-level analysis. I uploaded the con_*.img images and the design matrix could be created correctly. But when I run the estimate process it appears the following error:
> Error using gammaln
> Input must be nonnegative.
> Error in spm_ECdensity (line 70)
> EC(4,:) = a^(3/2)*exp(gammaln((v+k-3)/2)-b)...
> Error in spm_P_RF (line 63)
> EC = spm_ECdensity(STAT,Z,df);
> Error in spm_uc_peakFDR (line 83)
> [P,p,Eu] = spm_P_RF(1,0,ui,df,STAT,R,n);
> Error in spm_getSPM (line 761)
> [up,Pp] = spm_uc_peakFDR(0.05,df,STAT,R,n,Zum,XYZum,u);
> Error in spm_results_ui
> Error while evaluating UIControl Callback
> What is possible to do to fix it?
> Thanks in advance,
Guillaume Flandin, PhD
Wellcome Centre for Human Neuroimaging
University College London
12 Queen Square
London WC1N 3BG