While I am varying the parameters and runs the pcr file in fullprof I am observing the Singular matrix problem can anyone suggest me how to vary that parameter on which parameter this problem shown.
If I recall correctly, when you get the error FullProf should also tell you which parameter it is having issues with. Could you check if this is the case? In general, it probably means that you either have your scale factor set too low (or some other parameter set to zero that shouldn't be).
From a Mathematical point of view singular Matrix means that the correlation between two or more refined parameters is 1 (or -1), i.e. they decribe the same quantity. You should consider the correlation Matrix to see where is the mistake.
Worng guess value of refined parameters may cause it. Try with suitable value of the corresponding parameters.
Thanks to all of your helpful answers. I am trying with your suggestion and try to solve this problem Thanks for all of you.
In fact this problem is common for shape parameters, when you start to vary them from the values inserted in default. You can try to manually move them to get to a profile near the experimental one and only then let them to vary. In some cases this kind of error occurs when you leave occupancies and thermal parameters to vary together
Example if u refine Shift and Lattice parameter while both has 100% correlation.
That's why u have to follow the refinement strategy to avoid this.
there are no specific reason for such a problem, you will get it when some parameters conflict with each other on there is a wrong value which may lead to uncountable equation inside the software. any how it would be better to recreate your PCR file and start from the beginning and to take a backup after another when you doing the refinement.
you can also trace the conflict reason somehow by editing the pcr file using the note bad or any text editor available, you would notice that the region of error are zeroes or special character
In such case, you should be looking for higher order background polynomial and refine that only to fit up your profile, another reason for the error might originate from implementing the wrong space group, look for the correct one and then refine your parameters to avoid singular matrix error.
First check what is the values of U, V and W. they must be less than 0.5
Singular Matrix Error" conv. not yet reached ->[Max] Shift (Bover_ph1_pat1)/(eps*Sigma) = 147.45> 1======================>>>CYCLE:
anyone can help me in this?
- Badges
- Science topic
- Similar topics
- Filing