Hi,
I found a problem in the MPI version of CtfRefine (and several others)
that error messages are not correctly displayed. The latest commit
fixes this problem.
Best regards,
Takanori Nakane
> Dear Takanori,
>
> Thanks for the tip, I will try to find the culprit.
>
> Perhaps it may be possible in future to modify code such that error
> message lists problematic file?
>
> Thanks,
>
> Leo
>
> Prof. Leonid Sazanov
> IST Austria
> Am Campus 1
> A-3400 Klosterneuburg
> Austria
>
> Phone: +43 2243 9000 3026
> E-mail: [log in to unmask]
>
> On 09/01/19 13:43, Takanori Nakane wrote:
>> Hi,
>>
>> Probably one of your MRCS files is truncated; for example,
>> your STAR file points to the 200-th image in the stack, but the
>> MRCS file contains only 190.
>>
>> Best regards,
>>
>> Takanori Nakane
>>
>>> Dear Sjors,
>>>
>>> As I mentioned, error messages looked a bit different each time, so I
>>> did not post them initially.
>>>
>>> Now I noticed that all of them eventually trace back to the same line
>>> in
>>> code, related to MRC file read/write:
>>>
>>> in: /localhome/schloegl/src/relion-3.0-beta/src/rwMRC.h, line 191
>>> terminate called after throwing an instance of 'RelionError'
>>>
>>> This is the example command:
>>>
>>> `which relion_ctf_refine` --i Refine3D/job184/run_data.star --f
>>> PostProcess/job196/postprocess.star --m1
>>> Refine3D/job184/run_half1_class001_unfil.mrc --m2
>>> Refine3D/job184/run_half2_class001_unfil.mrc --mask
>>> MaskCreate/job156/mask_156.mrc --o CtfRefine/job226/ --kmin_defocus 30
>>> --fit_defocus --range 2000 --fit_beamtilt --kmin_tilt 30
>>> --only_do_unfinished? --j 8
>>>
>>> With this set of particles (~30K, ~3.8 A resolution) CtfRefine always
>>> fails, but with another set from the same micrographs (170K, 3.9 A
>>> resolution) the same type of command works.
>>>
>>> I double checked that all files are in place and there are no issues
>>> with permissions, so I am not sure why MRC file read/writing fails.
>>>
>>> Thanks,
>>>
>>> Leo
>>>
>>> Prof. Leonid Sazanov
>>> IST Austria
>>> Am Campus 1
>>> A-3400 Klosterneuburg
>>> Austria
>>>
>>> Phone: +43 2243 9000 3026
>>> E-mail: [log in to unmask]
>>>
>>> On 07/01/19 09:04, Sjors Scheres wrote:
>>>> Dear Leo,
>>>>
>>>> I'm afraid this message does not provide a good starting point for a
>>>> follow-up investigation on our side. The below is not only for you,
>>>> but
>>>> also for the community.
>>>>
>>>> When reporting errors, it is very useful for us to:
>>>> 1) know the exact command line arguments
>>>> 2) know the exact error
>>>> 3) know when that error occurs and when not
>>>> 4) (most useful) have a small test case you could share with us that
>>>> quickly and reproducibly leads to the error.
>>>>
>>>> In such cases we can relatively quickly find the error. Otherwise, it
>>>> may
>>>> take a long time...
>>>>
>>>> HTH,
>>>> Sjors
>>>>
>>>>
>>>>> Hi, recently using Relion3 build from 04 Dec with CtfRefine I
>>>>> constantly
>>>>> get cryptic errors which are different depending on whether I use
>>>>> several
>>>>> mpi and several threads or one mpi and several threads or even if
>>>>> just
>>>>> 1
>>>>> mpi and 1 thread. Larger set of particles from the same set of images
>>>>> works fine, so that is really puzzling.
>>>>> Any info on what may help highly appreciated.
>>>>> Thanks!
>>>>>
>>>>> ########################################################################
>>>>>
>>>>> To unsubscribe from the CCPEM list, click the following link:
>>>>> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1
>>>>>
>>> ########################################################################
>>>
>>> To unsubscribe from the CCPEM list, click the following link:
>>> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1
>>>
>>
>
> ########################################################################
>
> To unsubscribe from the CCPEM list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1
>
########################################################################
To unsubscribe from the CCPEM list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1
|