Hi,
I have gone through the tutorial and the mailing list archive but I was unable to find whether it is possible to measure the correlation only between some sites of the lattice under study, say about O(n) of possible correaltions not every possible 2 sites in the lattice which is about O(n^2) and is definitely a waste of calculations in my case. So, is there any way that I can give a list (of two sites) to the program? If not, can you think of a very easy workaround?
My main interest right now is the new DMRG MPS code, so even if there is any MPS specific technique that I can use I would be happy to hear about it.
Bests, Amir
Dear Amir,
Yes it is possible. Here you find the list of all parameters. http://alps.comp-phys.org/static/mps_doc/parameters.html http://alps.comp-phys.org/static/mps_doc/parameters.html
What you can use is: 1) MEASURE_HALF_CORRELATIONS[Name] which measures all i<j. So just N*N/2 calculations. 2) MEASURE_LOCAL_AT where you can specify at which position you want to measure the observables. For example MEASURE_LOCAL_AT[My Correlation] = “cdag_up:c_up|(0,1),(0,2),(0,3),(0,4)” to measure at i=0 and j=1,2,3,4. Note that this kind of measurements won’t be super-fast.
Best regards, Michele
-- ETH Zurich Michele Dolfi Institute for Theoretical Physics HIT G 32.4 Wolfgang-Pauli-Str. 27 8093 Zurich Switzerland
dolfim@phys.ethz.ch www.itp.phys.ethz.ch
+41 44 633 78 56 phone +41 44 633 11 15 fax
On 09 Mar 2015, at 18:19, Amir M.Aghaei amir.maghaei@gmail.com wrote:
Hi,
I have gone through the tutorial and the mailing list archive but I was unable to find whether it is possible to measure the correlation only between some sites of the lattice under study, say about O(n) of possible correaltions not every possible 2 sites in the lattice which is about O(n^2) and is definitely a waste of calculations in my case. So, is there any way that I can give a list (of two sites) to the program? If not, can you think of a very easy workaround?
My main interest right now is the new DMRG MPS code, so even if there is any MPS specific technique that I can use I would be happy to hear about it.
Bests, Amir
Dear Michele,
Thank for your prompt response. The second one looks like what I am looking for; however what do you mean by it is not superfast? for example, if I am interested on a nearest neighbor correlation over all of a lattice, which is O(n) correlations, will you recommend this option or the default given correaltion measurement (O(n^2))?
Bests, Amir
On Mon, Mar 9, 2015 at 1:57 PM, Michele Dolfi dolfim@phys.ethz.ch wrote:
Dear Amir,
Yes it is possible. Here you find the list of all parameters. http://alps.comp-phys.org/static/mps_doc/parameters.html
What you can use is:
- MEASURE_HALF_CORRELATIONS[Name] which measures all i<j. So just N*N/2
calculations. 2) MEASURE_LOCAL_AT where you can specify at which position you want to measure the observables. For example MEASURE_LOCAL_AT[My Correlation] = "cdag_up:c_up|(0,1),(0,2),(0,3),(0,4)" to measure at i=0 and j=1,2,3,4. Note that this kind of measurements won't be super-fast.
Best regards, Michele
-- ETH Zurich Michele Dolfi Institute for Theoretical Physics HIT G 32.4 Wolfgang-Pauli-Str. 27 8093 Zurich Switzerland
dolfim@phys.ethz.ch www.itp.phys.ethz.ch
+41 44 633 78 56 phone +41 44 633 11 15 fax
On 09 Mar 2015, at 18:19, Amir M.Aghaei amir.maghaei@gmail.com wrote:
Hi,
I have gone through the tutorial and the mailing list archive but I was unable to find whether it is possible to measure the correlation only between some sites of the lattice under study, say about O(n) of possible correaltions not every possible 2 sites in the lattice which is about O(n^2) and is definitely a waste of calculations in my case. So, is there any way that I can give a list (of two sites) to the program? If not, can you think of a very easy workaround?
My main interest right now is the new DMRG MPS code, so even if there is any MPS specific technique that I can use I would be happy to hear about it.
Bests, Amir
MEASURE_HALF_CORRELATIONS profits more from OpenMP parallelization, and it performs less contractions of the MPS network. I would anyway just try both.
Michele
-- ETH Zurich Michele Dolfi Institute for Theoretical Physics HIT G 32.4 Wolfgang-Pauli-Str. 27 8093 Zurich Switzerland
dolfim@phys.ethz.ch www.itp.phys.ethz.ch
+41 44 633 78 56 phone +41 44 633 11 15 fax
On 09 Mar 2015, at 22:03, Amir M.Aghaei amir.maghaei@gmail.com wrote:
Dear Michele,
Thank for your prompt response. The second one looks like what I am looking for; however what do you mean by it is not superfast? for example, if I am interested on a nearest neighbor correlation over all of a lattice, which is O(n) correlations, will you recommend this option or the default given correaltion measurement (O(n^2))?
Bests, Amir
On Mon, Mar 9, 2015 at 1:57 PM, Michele Dolfi <dolfim@phys.ethz.ch mailto:dolfim@phys.ethz.ch> wrote: Dear Amir,
Yes it is possible. Here you find the list of all parameters. http://alps.comp-phys.org/static/mps_doc/parameters.html http://alps.comp-phys.org/static/mps_doc/parameters.html
What you can use is:
- MEASURE_HALF_CORRELATIONS[Name] which measures all i<j. So just N*N/2 calculations.
- MEASURE_LOCAL_AT where you can specify at which position you want to measure the observables. For example MEASURE_LOCAL_AT[My Correlation] = “cdag_up:c_up|(0,1),(0,2),(0,3),(0,4)” to measure at i=0 and j=1,2,3,4. Note that this kind of measurements won’t be super-fast.
Best regards, Michele
-- ETH Zurich Michele Dolfi Institute for Theoretical Physics HIT G 32.4 Wolfgang-Pauli-Str. 27 8093 Zurich Switzerland
dolfim@phys.ethz.ch mailto:dolfim@phys.ethz.ch www.itp.phys.ethz.ch http://www.itp.phys.ethz.ch/
+41 44 633 78 56 tel:%2B41%2044%20633%2078%2056 phone +41 44 633 11 15 tel:%2B41%2044%20633%2011%2015 fax
On 09 Mar 2015, at 18:19, Amir M.Aghaei <amir.maghaei@gmail.com mailto:amir.maghaei@gmail.com> wrote:
Hi,
I have gone through the tutorial and the mailing list archive but I was unable to find whether it is possible to measure the correlation only between some sites of the lattice under study, say about O(n) of possible correaltions not every possible 2 sites in the lattice which is about O(n^2) and is definitely a waste of calculations in my case. So, is there any way that I can give a list (of two sites) to the program? If not, can you think of a very easy workaround?
My main interest right now is the new DMRG MPS code, so even if there is any MPS specific technique that I can use I would be happy to hear about it.
Bests, Amir
comp-phys-alps-users@lists.phys.ethz.ch