Dear Alps users,
First of all I want to congratulate you for the wonderful job of Alps.
I was looking for information, to check if Alps is able to apply Quantum Monte Carlo to fermion Hubbard Model in half filling. If right now is not able do you think that this is an feature that we could expect from Alps in the future?
Thanks in advance
João Lopes
Dear all,
I diagonalized a 2x2 spin 1/2 system with fulldiag for different magnetic fields and I got only 14 energies in my datasets. I would have expected 16. What is wrong ?
Many thanks,
Jean-David
On 1 Sep 2010, at 11:14, Jean-David Picon wrote:
Dear all,
I diagonalized a 2x2 spin 1/2 system with fulldiag for different magnetic fields and I got only 14 energies in my datasets. I would have expected 16. What is wrong ?
Many thanks,
Jean-David
Dear Jean-David,
can you please post the input files?
Matthias
Sorry,
Here is the python script I used to make the ED. By the way, as you can see, I also ask in the script for the Jz(i).Jz(j) correlations as in the demo script but the resulting datasets only have 2 components : the energy and the off-diagonal correlations (that I also want). Why is it so ?
Thanks,
Jean-David
On 09/01/2010 03:35 PM, Matthias Troyer wrote:
On 1 Sep 2010, at 11:14, Jean-David Picon wrote:
Dear all,
I diagonalized a 2x2 spin 1/2 system with fulldiag for different magnetic fields and I got only 14 energies in my datasets. I would have expected 16. What is wrong ?
Many thanks,
Jean-David
Dear Jean-David,
can you please post the input files?
Matthias
Dear Jean-David,
I get 16 energies in 14 symmetry sectors. The k=(0,0) Sz=0 sector has 3 states, all others just one. I will look into the missing diagonal spin correlations tomorrow.
Matthias
On 2 Sep 2010, at 16:34, Jean-David Picon wrote:
Sorry,
Here is the python script I used to make the ED. By the way, as you can see, I also ask in the script for the Jz(i).Jz(j) correlations as in the demo script but the resulting datasets only have 2 components : the energy and the off-diagonal correlations (that I also want). Why is it so ?
Thanks,
Jean-David
On 09/01/2010 03:35 PM, Matthias Troyer wrote:
On 1 Sep 2010, at 11:14, Jean-David Picon wrote:
Dear all,
I diagonalized a 2x2 spin 1/2 system with fulldiag for different magnetic fields and I got only 14 energies in my datasets. I would have expected 16. What is wrong ?
Many thanks,
Jean-David
Dear Jean-David,
can you please post the input files?
Matthias
<S_ed_Tommaso2.py>
Hi Jean-David,
You wrote an extra= in
'MEASURE_CORRELATIONS[Diagonal spin correlations]=' : 'Sz'
That means that this does not match the pattern for a correlation definition
Matthias
On 2 Sep 2010, at 16:34, Jean-David Picon wrote:
Sorry,
Here is the python script I used to make the ED. By the way, as you can see, I also ask in the script for the Jz(i).Jz(j) correlations as in the demo script but the resulting datasets only have 2 components : the energy and the off-diagonal correlations (that I also want). Why is it so ?
Thanks,
Jean-David
On 09/01/2010 03:35 PM, Matthias Troyer wrote:
On 1 Sep 2010, at 11:14, Jean-David Picon wrote:
Dear all,
I diagonalized a 2x2 spin 1/2 system with fulldiag for different magnetic fields and I got only 14 energies in my datasets. I would have expected 16. What is wrong ?
Many thanks,
Jean-David
Dear Jean-David,
can you please post the input files?
Matthias
<S_ed_Tommaso2.py>
Many thanks for both answers. It was actually not clear to me how to extract the degeneracy of each sector ....
Another question then : when I have a sector (Sz,k) with degeneracy 3, what is the exact meaning of the correlations S+(i)S-(j) on this sector ? More precisely, how do the 3 states contribute to the correlations ?
Thanks again.
Jean-David
On 09/02/2010 04:19 PM, Matthias Troyer wrote:
Hi Jean-David,
You wrote an extra= in
'MEASURE_CORRELATIONS[Diagonal spin correlations]=' : 'Sz'
That means that this does not match the pattern for a correlation definition
Matthias
On 2 Sep 2010, at 16:34, Jean-David Picon wrote:
Sorry,
Here is the python script I used to make the ED. By the way, as you can see, I also ask in the script for the Jz(i).Jz(j) correlations as in the demo script but the resulting datasets only have 2 components : the energy and the off-diagonal correlations (that I also want). Why is it so ?
Thanks,
Jean-David
On 09/01/2010 03:35 PM, Matthias Troyer wrote:
On 1 Sep 2010, at 11:14, Jean-David Picon wrote:
Dear all,
I diagonalized a 2x2 spin 1/2 system with fulldiag for different magnetic fields and I got only 14 energies in my datasets. I would have expected 16. What is wrong ?
Many thanks,
Jean-David
Dear Jean-David,
can you please post the input files?
Matthias
<S_ed_Tommaso2.py>
On 2 Sep 2010, at 23:14, Jean-David Picon wrote:
Many thanks for both answers. It was actually not clear to me how to extract the degeneracy of each sector ....
Another question then : when I have a sector (Sz,k) with degeneracy 3,
It's not degeneracy. There are three states with different energies.
what is the exact meaning of the correlations S+(i)S-(j) on this sector ? More precisely, how do the 3 states contribute to the correlations ?
The correlations are measured for each of the states and stored separately. Loading them gives a 2D array
Matthias
Ok ! Thanks.
Then in order to be complete, the tutorial tutorial1a.py http://alps.comp-phys.org/static/tutorials2.0.0/ed-01-sparsediag/tutorial1a.py should maybe be slightly modified because the current version only displays one measurement of each observable per (Sz,k) sector even if there are many.
Adding one line
for sector in data[0]: print '\nSector with Sz =', sector[0].props['Sz'], print 'and k =', sector[0].props['TOTAL_MOMENTUM'] for s in sector: * for i in range(pyalps.size(s.y)):* if pyalps.size(s.y[i])==1: print s.props['observable'], ' : ', s.y[i] else: for (x,y) in zip(s.x,s.y[i]): print s.props['observable'], '(', x, ') : ', y
corrects this point even if then the order of printing is not totally satisfactory.
Jean-David
On 09/02/2010 05:24 PM, Matthias Troyer wrote:
On 2 Sep 2010, at 23:14, Jean-David Picon wrote:
Many thanks for both answers. It was actually not clear to me how to extract the degeneracy of each sector ....
Another question then : when I have a sector (Sz,k) with degeneracy 3,
It's not degeneracy. There are three states with different energies.
what is the exact meaning of the correlations S+(i)S-(j) on this sector ? More precisely, how do the 3 states contribute to the correlations ?
The correlations are measured for each of the states and stored separately. Loading them gives a 2D array
Matthias
Do you want to go ahead and do that?
Matthias
Sent from my iPad
On Sep 3, 2010, at 17:32, Jean-David Picon jean-david.picon@epfl.ch wrote:
Ok ! Thanks.
Then in order to be complete, the tutorial tutorial1a.py should maybe be slightly modified because the current version only displays one measurement of each observable per (Sz,k) sector even if there are many.
Adding one line
for sector in data[0]: print '\nSector with Sz =', sector[0].props['Sz'], print 'and k =', sector[0].props['TOTAL_MOMENTUM'] for s in sector: for i in range(pyalps.size(s.y)): if pyalps.size(s.y[i])==1: print s.props['observable'], ' : ', s.y[i] else: for (x,y) in zip(s.x,s.y[i]): print s.props['observable'], '(', x, ') : ', y
corrects this point even if then the order of printing is not totally satisfactory.
Jean-David
On 09/02/2010 05:24 PM, Matthias Troyer wrote:
On 2 Sep 2010, at 23:14, Jean-David Picon wrote:
Many thanks for both answers. It was actually not clear to me how to extract the degeneracy of each sector ....
Another question then : when I have a sector (Sz,k) with degeneracy 3,
It's not degeneracy. There are three states with different energies.
what is the exact meaning of the correlations S+(i)S-(j) on this sector ? More precisely, how do the 3 states contribute to the correlations ?
The correlations are measured for each of the states and stored separately. Loading them gives a 2D array
Matthias
Dear João
We have plans to include cluster DMFT schemes in ALPS but not a determinental QMC code for just the half filled Hubbaf model since it is of too limited use. However I can get you in touch with people who have such a code.
Matthias
Sent from my iPad
On Sep 1, 2010, at 2:03, João Manuel Viana Parente Lopes jvianalopes@gmail.com wrote:
Dear Alps users,
First of all I want to congratulate you for the wonderful job of Alps.
I was looking for information, to check if Alps is able to apply Quantum Monte Carlo to fermion Hubbard Model in half filling. If right now is not able do you think that this is an feature that we could expect from Alps in the future?
Thanks in advance
João Lopes
comp-phys-alps-users@lists.phys.ethz.ch