search

TRIQS / triqs_0.x

DEPRECATED -- This is the repository of the older versions of TRIQS
Other
11 stars 9 forks source link

momentum-resolved spectral function problems. #85

Closed blueway39 closed 11 years ago

blueway39 commented 12 years ago

Hello all.

To produce a momentum-resolved spectral function, I did as 'Introduction to dmftproj' and web.

x lapw1 -band (done)

x lapw2 -alm -band (error)

Error in LAPW2 'FERMI' - number of k-points inconsistent when reading kgen 'FERMI' - check IN1 and KGEN files!

But, instead of 'x lapw2 -alm -band', 'x lapw2 -qtl -band' is done well and 'dmftproj -band' is done too.

Is it right way?

And I made 'band.py' for calculating the momentum-resolved spectral function.

There is no example, so I'm not sure whether it is right script.

The script is on web : https://github.com/blueway39/all/issues/2

When I use this script, I got error message.

$ pytriqs band.py

Starting on 1 Nodes at : 2012-10-14 17:34:39.123203 Repacking the file SrVO3_01.h5 Reading input from SrVO3_01.ctqmcout... Reading symmetry input from SrVO3_01.symqmc... Reading bands input from SrVO3_01.outband... Traceback (most recent call last): File "test.py", line 28, in Converter.convert_bands_input() File "/opt/Triqs/TRIQS/lib/python2.7/site-packages/pytriqs/Wien2k/SumK_LDA_Wien2k_input.py", line 438, in convert_bands_input for it in thingstowrite: exec "ar['%s']['%s'] = %s"%(self.BandsSubGrp,it,it) File "", line 1, in File "/opt/Triqs/TRIQS/lib/python2.7/site-packages/pytriqs/Base/Archive/HDF_Archive.py", line 203, in setitem for n,v in d.items() : SUB[n] = v File "/opt/Triqs/TRIQS/lib/python2.7/site-packages/pytriqs/Base/Archive/HDF_Archive.py", line 203, in setitem for n,v in d.items() : SUB[n] = v File "/opt/Triqs/TRIQS/lib/python2.7/site-packages/pytriqs/Base/Archive/HDF_Archive.py", line 203, in setitem for n,v in d.items() : SUB[n] = v File "/opt/Triqs/TRIQS/lib/python2.7/site-packages/pytriqs/Base/Archive/HDF_Archive.py", line 201, in setitem if (d=={}) : raise ValueError, "reduce_to_dict returns an empty dict" ValueError: reduce_to_dict returns an empty dict

Three questions.

  1. Is it right way that I use 'x lapw2 -qtl -band' instead of 'x lapw2 -alm -band'?
  2. My script for calculating the momentum-resolved spectral function can work?
  3. Why is the error occurred?

Thank you for your attention.

leopo commented 12 years ago

I believe the flag -alm has now been changed to -almd, therefore, the correct commands are:

x lapw1 -band

x lapw2 -almd -band

Please try this

blueway39 commented 12 years ago

'x lapw2 -almd -band' has same error.

Error in LAPW2 'FERMI' - number of k-points inconsistent when reading kgen 'FERMI' - check IN1 and KGEN files!

aichhorn commented 12 years ago

Okay, this looks like a wien2k problem and not a triqs problem. lapw2 does not like the k-point list that you generated. what happens if you calculate the band structure (i.e. spaghetti) in wien2k, without triqs/dmft? does that work? and x lapw2 -almd without the -band flag does work?

blueway39 commented 12 years ago

The calculation of band structure in Wien2k(only LDA results) is working.

But, 'x lapw2 -almd' returns error.

aichhorn commented 12 years ago

What version of Wien2k are you using?

blueway39 commented 12 years ago

I use Wien2k(ver. 11.1).

aichhorn commented 12 years ago

Okay, just to be sure: 'x lapw2 -almd' and 'x lapw2 -almd -band' do not work but crash with the Error message as given above. Or is it just the -band option that causes problems (This thread is really messy)? I guess that your problem comes from inconsistencies of .klist files. You can check also on the Wien2k mailing list for this error. It is a Wien2k problem in the end. Or restart with a clean and empty directory, restore a converged Wien2k run, construct the klist_band file, and then do 'x lapw1 -band' and 'x lapw2 -almd -band'.