The meaning of the 'TruncatedWeight' in ALPS MPS simulation
Dear ALPS users, The ALPS MPS code can print 'TruncatedWeight' to estimate the error of the simulation. I have a question: What is the meaning of this quantity and the relation with the truncation error in traditional DMRG simulation? For Fig. 5 in Computer Physics Communications, 185(12):3430 – 3440, 2014., the quantity may be summed as the final truncation error after a sweep. Thanks a lot! Best regards, Rongyang Sun
Dear Rongyang, The TruncatedWeight is the sum of the square of the discarded singular values during a truncation. This is the same value used in other DMRG applications. Best regards, Michele
On 7 Jan 2018, at 08:00, Rongyang Sun <sun-rongyang@outlook.com> wrote:
Dear ALPS users,
The ALPS MPS code can print 'TruncatedWeight' to estimate the error of the simulation. I have a question:
What is the meaning of this quantity and the relation with the truncation error in traditional DMRG simulation?
For Fig. 5 in Computer Physics Communications, 185(12):3430 – 3440, 2014., the quantity may be summed as the final truncation error after a sweep.
Thanks a lot!
Best regards, Rongyang Sun
---- Comp-phys-alps-users Mailing List for the ALPS Project http://alps.comp-phys.org/ <http://alps.comp-phys.org/>
List info: https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users <https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users> Archive: https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users <https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users>
Unsubscribe by writing a mail to comp-phys-alps-users-leave@lists.phys.ethz.ch <mailto:comp-phys-alps-users-leave@lists.phys.ethz.ch>.
Dear Michele, Thanks so much for your response. I got a very strange truncation error pattern which is attached. What is the meaning of this kind of pattern? The total energy of the simulation matches the other DMRG result for the same model parameters to 1e-5. The truncation error is calculated by the same method in Computer Physics Communications, 185(12):3430 – 3440, 2014. Thanks. Best regards, Rongyang ________________________________ From: Comp-phys-alps-users <comp-phys-alps-users-bounces@lists.phys.ethz.ch> on behalf of Michele Dolfi <michele.dolfi@gmail.com> Sent: Sunday, January 7, 2018 3:45:41 PM To: comp-phys-alps-users@lists.phys.ethz.ch Subject: Re: [ALPS-users] The meaning of the 'TruncatedWeight' in ALPS MPS simulation Dear Rongyang, The TruncatedWeight is the sum of the square of the discarded singular values during a truncation. This is the same value used in other DMRG applications. Best regards, Michele On 7 Jan 2018, at 08:00, Rongyang Sun <sun-rongyang@outlook.com<mailto:sun-rongyang@outlook.com>> wrote: Dear ALPS users, The ALPS MPS code can print 'TruncatedWeight' to estimate the error of the simulation. I have a question: What is the meaning of this quantity and the relation with the truncation error in traditional DMRG simulation? For Fig. 5 in Computer Physics Communications, 185(12):3430 – 3440, 2014., the quantity may be summed as the final truncation error after a sweep. Thanks a lot! Best regards, Rongyang Sun ---- Comp-phys-alps-users Mailing List for the ALPS Project http://alps.comp-phys.org/ List info: https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users Archive: https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users Unsubscribe by writing a mail to comp-phys-alps-users-leave@lists.phys.ethz.ch<mailto:comp-phys-alps-users-leave@lists.phys.ethz.ch>.
Hello, Well, the figure seems to indicate that something is still changing a bit in the wavefunction. I think it would be interesting to see what happens by performing more sweeps, and by looking at the convergence of other observables. In general it is important to note that the energy converges faster than other observables, e.g. long-range quantities. A typical example is: http://dx.doi.org/10.1103/PhysRevLett.115.056401 <http://dx.doi.org/10.1103/PhysRevLett.115.056401> Best, Michele
On 10 Jan 2018, at 02:33, Rongyang Sun <sun-rongyang@outlook.com> wrote:
Dear Michele,
Thanks so much for your response. I got a very strange truncation error pattern which is attached. What is the meaning of this kind of pattern? The total energy of the simulation matches the other DMRG result for the same model parameters to 1e-5. The truncation error is calculated by the same method in Computer Physics Communications, 185(12):3430 – 3440, 2014. Thanks.
Best regards, Rongyang From: Comp-phys-alps-users <comp-phys-alps-users-bounces@lists.phys.ethz.ch> on behalf of Michele Dolfi <michele.dolfi@gmail.com> Sent: Sunday, January 7, 2018 3:45:41 PM To: comp-phys-alps-users@lists.phys.ethz.ch Subject: Re: [ALPS-users] The meaning of the 'TruncatedWeight' in ALPS MPS simulation
Dear Rongyang,
The TruncatedWeight is the sum of the square of the discarded singular values during a truncation. This is the same value used in other DMRG applications.
Best regards, Michele
On 7 Jan 2018, at 08:00, Rongyang Sun <sun-rongyang@outlook.com <mailto:sun-rongyang@outlook.com>> wrote:
Dear ALPS users,
The ALPS MPS code can print 'TruncatedWeight' to estimate the error of the simulation. I have a question:
What is the meaning of this quantity and the relation with the truncation error in traditional DMRG simulation?
For Fig. 5 in Computer Physics Communications, 185(12):3430 – 3440, 2014., the quantity may be summed as the final truncation error after a sweep.
Thanks a lot!
Best regards, Rongyang Sun
---- Comp-phys-alps-users Mailing List for the ALPS Project http://alps.comp-phys.org/ <http://alps.comp-phys.org/>
List info: https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users <https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users> Archive: https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users <https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users>
Unsubscribe by writing a mail to comp-phys-alps-users-leave@lists.phys.ethz.ch <mailto:comp-phys-alps-users-leave@lists.phys.ethz.ch>.
<truncationerror.pdf>
---- Comp-phys-alps-users Mailing List for the ALPS Project http://alps.comp-phys.org/ <http://alps.comp-phys.org/>
List info: https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users <https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users> Archive: https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users <https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users>
Unsubscribe by writing a mail to comp-phys-alps-users-leave@lists.phys.ethz.ch <mailto:comp-phys-alps-users-leave@lists.phys.ethz.ch>.
Dear Michele, Thanks for your answer with patience. I will check my simulation results more carefully. Best, Rongyang ________________________________ From: Comp-phys-alps-users <comp-phys-alps-users-bounces@lists.phys.ethz.ch> on behalf of Michele Dolfi <michele.dolfi@gmail.com> Sent: Thursday, January 11, 2018 4:14:44 AM To: comp-phys-alps-users@lists.phys.ethz.ch Subject: Re: [ALPS-users] The meaning of the 'TruncatedWeight' in ALPS MPS simulation Hello, Well, the figure seems to indicate that something is still changing a bit in the wavefunction. I think it would be interesting to see what happens by performing more sweeps, and by looking at the convergence of other observables. In general it is important to note that the energy converges faster than other observables, e.g. long-range quantities. A typical example is: http://dx.doi.org/10.1103/PhysRevLett.115.056401 Best, Michele On 10 Jan 2018, at 02:33, Rongyang Sun <sun-rongyang@outlook.com<mailto:sun-rongyang@outlook.com>> wrote: Dear Michele, Thanks so much for your response. I got a very strange truncation error pattern which is attached. What is the meaning of this kind of pattern? The total energy of the simulation matches the other DMRG result for the same model parameters to 1e-5. The truncation error is calculated by the same method in Computer Physics Communications, 185(12):3430 – 3440, 2014. Thanks. Best regards, Rongyang ________________________________ From: Comp-phys-alps-users <comp-phys-alps-users-bounces@lists.phys.ethz.ch<mailto:comp-phys-alps-users-bounces@lists.phys.ethz.ch>> on behalf of Michele Dolfi <michele.dolfi@gmail.com<mailto:michele.dolfi@gmail.com>> Sent: Sunday, January 7, 2018 3:45:41 PM To: comp-phys-alps-users@lists.phys.ethz.ch<mailto:comp-phys-alps-users@lists.phys.ethz.ch> Subject: Re: [ALPS-users] The meaning of the 'TruncatedWeight' in ALPS MPS simulation Dear Rongyang, The TruncatedWeight is the sum of the square of the discarded singular values during a truncation. This is the same value used in other DMRG applications. Best regards, Michele On 7 Jan 2018, at 08:00, Rongyang Sun <sun-rongyang@outlook.com<mailto:sun-rongyang@outlook.com>> wrote: Dear ALPS users, The ALPS MPS code can print 'TruncatedWeight' to estimate the error of the simulation. I have a question: What is the meaning of this quantity and the relation with the truncation error in traditional DMRG simulation? For Fig. 5 in Computer Physics Communications, 185(12):3430 – 3440, 2014., the quantity may be summed as the final truncation error after a sweep. Thanks a lot! Best regards, Rongyang Sun ---- Comp-phys-alps-users Mailing List for the ALPS Project http://alps.comp-phys.org/ List info: https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users Archive: https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users Unsubscribe by writing a mail to comp-phys-alps-users-leave@lists.phys.ethz.ch<mailto:comp-phys-alps-users-leave@lists.phys.ethz.ch>. <truncationerror.pdf> ---- Comp-phys-alps-users Mailing List for the ALPS Project http://alps.comp-phys.org/ List info: https://lists.phys.ethz.ch//listinfo/comp-phys-alps-users Archive: https://lists.phys.ethz.ch//pipermail/comp-phys-alps-users Unsubscribe by writing a mail to comp-phys-alps-users-leave@lists.phys.ethz.ch<mailto:comp-phys-alps-users-leave@lists.phys.ethz.ch>.
participants (2)
-
Michele Dolfi
-
Rongyang Sun