Hi Huy,
Sometimes zero-sum-of-weight errors are caused because the
translational/rotational searches are too coarse for the prior chosen. By
default translational priors have a sigma of 3 pixels. Perhaps you used a
much coarser translational sampling than that?
Otherwise, indeed sometimes it's problems with 'strange particles'
HTH,
Sjors
> Dear Sjors,
>
> I tried 2D classification of my data, I got the Zero sum of weight error
> (At the end of the email) after noise estimation. The 3D classification of
> the same data is fine.
>
>>From reading in the mailing list, the zero sum of weight is because of
>> all
> black image or image with extreme value. I checked and I don't have any
> black image. Is there any way for me to identify the troubling image
> (part_id) with the image name in the star file so that I can remove them?
>
> Best regards,
> Huy
>
> Error msg here:
> In thread 1
> exp_thisparticle_sumweight= 0
> part_id= 126646
> group_id= 2973 mymodel.scale_correction[group_id]= 1
> exp_ipass= 0
> sampling.NrDirections(0, true)= 1 sampling.NrDirections(0, false)= 1
> sampling.NrPsiSamplings(0, true)= 36 sampling.NrPsiSamplings(0, false)=
> 36
> mymodel.sigma2_noise[ipart]=
> 6.1e-16
> .....
> wsum_model.pdf_direction[ipart]=
> 123.87828
>
> mymodel.avg_norm_correction= 1
> wsum_model.avg_norm_correction= 89.22
> written out Mweight.spi
> exp_thisparticle_sumweight= 0
> exp_min_diff2[ipart]= 881.537
> ERROR!!! zero sum of weights....
> File: src/ml_optimiser.cpp line: 4103
>
--
Sjors Scheres
MRC Laboratory of Molecular Biology
Francis Crick Avenue, Cambridge Biomedical Campus
Cambridge CB2 0QH, U.K.
tel: +44 (0)1223 267061
http://www2.mrc-lmb.cam.ac.uk/groups/scheres
|