ALF issueshttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues2021-06-07T12:15:42Zhttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/192More compilers2021-06-07T12:15:42ZFakher F. AssaadMore compilersWe would need to get our hands on a Cray, IBM and NAG fortran compilers to test ALF. ToDo see which computer centers in Germany would have them?We would need to get our hands on a Cray, IBM and NAG fortran compilers to test ALF. ToDo see which computer centers in Germany would have them?Fakher F. AssaadFakher F. Assaadhttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/182Uninitialized variables in upgrade.F90 and in Mat_subroutines.F902021-04-12T11:37:22ZFakher F. AssaadUninitialized variables in upgrade.F90 and in Mat_subroutines.F90Warnings are produced with the -Wuninitialized flag for the above mentioned routines.Warnings are produced with the -Wuninitialized flag for the above mentioned routines.https://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/177FFT for application of Op_T matrices2021-03-24T17:53:33ZFlorian GothFFT for application of Op_T matricesFakher wants to use an FFT to optimize for circulant matrices.Fakher wants to use an FFT to optimize for circulant matrices.Fakher F. AssaadFakher F. Assaadhttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/176Block Diagonal Op_T2020-11-30T13:03:23ZFlorian GothBlock Diagonal Op_TFakher wants to automatically decompose block diagonal matrices.Fakher wants to automatically decompose block diagonal matrices.Fakher F. AssaadFakher F. Assaadhttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/175Include MSCBDECOMP + higher order checkerboard2021-01-06T01:03:41ZFlorian GothInclude MSCBDECOMP + higher order checkerboard- [x] include MSCBDECOMP
- [x] make an object for it
- [x] Higher order checkerboard- [x] include MSCBDECOMP
- [x] make an object for it
- [x] Higher order checkerboardFlorian GothFlorian Gothhttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/174More Use cases for the ContainerElementBase Object Hierarchy2020-12-03T15:16:37ZFlorian GothMore Use cases for the ContainerElementBase Object HierarchyNow with the ContainerElementBase Object Hierarchy in place and a dynamic container present that is able to take up
these objects in some "dynamic" way we can think about extensions.
## Approximate techniques
- [ ] Minimum Split Checker...Now with the ContainerElementBase Object Hierarchy in place and a dynamic container present that is able to take up
these objects in some "dynamic" way we can think about extensions.
## Approximate techniques
- [ ] Minimum Split Checkerboard + Higher Order Checkerboard #175
- [ ] Lanczos
## Exact techniques
- [ ] Block techniques #176
- [ ] FFT techniques
## Use cases in other places
- [ ] Op_V is not that much different. Get rid of the various types?Florian GothFlorian Gothhttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/170Joining all possible global updates options2020-11-09T18:27:10ZJefferson Stafusa E. PortelaJoining all possible global updates optionsLangevin and HMC are now options of Global_update_scheme - what about integrating into this variable also the custom/user-defined global update option (Global_moves, ? Global_tau_moves ?), to avoid having conceptually similar options sca...Langevin and HMC are now options of Global_update_scheme - what about integrating into this variable also the custom/user-defined global update option (Global_moves, ? Global_tau_moves ?), to avoid having conceptually similar options scattered into different parameters.Fakher F. AssaadFakher F. Assaad2020-11-30https://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/169Flo's playground with Op_Ts2020-11-30T16:15:54ZFlorian GothFlo's playground with Op_TsSome playground for ideas.Some playground for ideas.Florian GothFlorian Gothhttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/168Doc homogenization2020-11-02T12:29:51ZFlorian GothDoc homogenizationA place to collect things that need homogenization:
- SU(N) U(N), O(2N), etc. math or not math?
- Datatypes: INT, DBLE, REAL, CHAR, CMPLX
- Sentences in tables: capital 1st letter and full stop at the end or not?A place to collect things that need homogenization:
- SU(N) U(N), O(2N), etc. math or not math?
- Datatypes: INT, DBLE, REAL, CHAR, CMPLX
- Sentences in tables: capital 1st letter and full stop at the end or not?ALF-2.0https://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/167Why is UDV%D a full complex number2020-10-26T12:41:40ZFlorian GothWhy is UDV%D a full complex number@fassaad, @Hofmann :
UDV%D is still a full complex number. We calculate D from the return values of norm related functions (Abs) which give positive real numbers. To my knowledge we never store anything complex in this array anyway and h...@fassaad, @Hofmann :
UDV%D is still a full complex number. We calculate D from the return values of norm related functions (Abs) which give positive real numbers. To my knowledge we never store anything complex in this array anyway and hence we might want to optimize that away.Florian GothFlorian Gothhttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/163Max_Ent2020-09-27T09:56:58ZFakher F. AssaadMax_Ent@Jonas To make MaxEnt work better with ALF I have replaced
write(File_out,'(A,"_",F4.2,"_",F4.2,"/g_",F4.2,"_",F4.2)') trim(name_obs), Xk_p(1,n), Xk_p(2,n), Xk_p(1,n), Xk_p(2,n) with
write(File_out,'(A,"_",F4.2,"_",F4.2,"/g_dat")') tri...@Jonas To make MaxEnt work better with ALF I have replaced
write(File_out,'(A,"_",F4.2,"_",F4.2,"/g_",F4.2,"_",F4.2)') trim(name_obs), Xk_p(1,n), Xk_p(2,n), Xk_p(1,n), Xk_p(2,n) with
write(File_out,'(A,"_",F4.2,"_",F4.2,"/g_dat")') trim(name_obs), Xk_p(1,n), Xk_p(2,n)
and
write(File_out,'(A,"_R0/g_R0")') trim(name_obs) with
write(File_out,'(A,"_R0/g_dat")') trim(name_obs)
Now we do not have to move any files aside from the parameter file before Running MaxEnt.
I have also changed the doc accordingly.
Best Fakher.https://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/161Boundaries of meassurement interval2020-09-26T13:51:05ZJohannes HofmannBoundaries of meassurement intervalAdapt the interval in which measurements take place, LOBS_ST and LOBS_EN.Adapt the interval in which measurements take place, LOBS_ST and LOBS_EN.Johannes HofmannJohannes Hofmannhttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/159Documentation: Get rid of object names in tables detailing members of types2020-09-15T14:56:33ZJonas SchwabDocumentation: Get rid of object names in tables detailing members of typesMost of the tables in the documentation detailing the members of a type have an object name prepended, e.g. `Latt%N`. How about removing the prepended parts? Since saying they're part of a type and prepending an (arbitrary) instance name...Most of the tables in the documentation detailing the members of a type have an object name prepended, e.g. `Latt%N`. How about removing the prepended parts? Since saying they're part of a type and prepending an (arbitrary) instance name is a bit redundant. This way, we would also get some space in the often quite crowded tables.Jonas SchwabJonas Schwabhttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/152Dictionary of stopcodes2020-08-04T10:28:32ZJonas SchwabDictionary of stopcodesCreate a dictionary of exit codes, to signify which kind of error occurred.
- 1: Generic error
- ...Create a dictionary of exit codes, to signify which kind of error occurred.
- 1: Generic error
- ...https://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/150Job Failed #130624 : PGI19.10 miscompiles tests2020-12-05T12:35:35ZFlorian GothJob Failed #130624 : PGI19.10 miscompiles testsJob [#130624](https://git.physik.uni-wuerzburg.de/ALF/ALF/-/jobs/130624) failed for fa1f3cfebd4c94091e7f1afd9c9a538530a4def9:
The testsuite still runs successfully with PGI18.10.
PGI19.10 with -O3 as requested by the configure script m...Job [#130624](https://git.physik.uni-wuerzburg.de/ALF/ALF/-/jobs/130624) failed for fa1f3cfebd4c94091e7f1afd9c9a538530a4def9:
The testsuite still runs successfully with PGI18.10.
PGI19.10 with -O3 as requested by the configure script miscompiles some tests and hence the test suite runs fail.
Lowering this to -O1 fixes it.https://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/149Let configure.sh change Makefile2020-08-03T11:23:16ZJonas SchwabLet configure.sh change MakefileWith a few additional lines, one can let configure.sh write the environment variables to the Makefile, such that it doesn't need to be sourced any more and therefore won't change the environment of the user.With a few additional lines, one can let configure.sh write the environment variables to the Makefile, such that it doesn't need to be sourced any more and therefore won't change the environment of the user.https://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/144reweighting to solve large/infinite variance issues (spikes)2020-07-29T13:43:14ZJohannes Hofmannreweighting to solve large/infinite variance issues (spikes)I came across a model that seems to suffer from rare events due to configurations with almost vanishing weight but large contributions to the observables (energy).
The is a proposed rather generic algorithm (https://journals.aps.org/pre/...I came across a model that seems to suffer from rare events due to configurations with almost vanishing weight but large contributions to the observables (energy).
The is a proposed rather generic algorithm (https://journals.aps.org/pre/pdf/10.1103/PhysRevE.93.033303) in which one essentially samples 1 additional time slice, that is then ignored during the measurements. The idea is that tracing out the auxiliary fields on one time-slice (the one you are measuring on) guarantees a finite weight p(X) = sum_x p(X,x), given p(X,x) is non-negative, i.e. no sign-problem. One can also view this procedure as a reweighting scheme with <exp(-dtau H)>. I think this can be easily implemented for equal-time observables, potentially even on the Hamiltonian side with a bit of knowledge about the core routines. The time-displaced routines seem to be a bit trickier on first sight and might require some small changes in the core routines (I'm not sure yet).https://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/140Consistent nomenclature: Phi_X, Phi_Y or Phi_1, Phi_2?2020-07-21T13:36:24ZJefferson Stafusa E. PortelaConsistent nomenclature: Phi_X, Phi_Y or Phi_1, Phi_2?The code uses variables Phi_X and Phi_Y for the twist on the boundaries, but those directions (x and y) have otherwise been defined by the indices 1 and 2. We might want to make this consistent at some point.The code uses variables Phi_X and Phi_Y for the twist on the boundaries, but those directions (x and y) have otherwise been defined by the indices 1 and 2. We might want to make this consistent at some point.2020-09-30https://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/139analysis: connected correlations2023-03-13T13:10:23ZFrancesco Parisen Toldinanalysis: connected correlationsThe analysis program ```cov_eq``` computes a subtraction of the "disconnected" part of the correlations [here](https://git.physik.uni-wuerzburg.de/ALF/ALF/-/blob/master/Analysis/cov_eq.F90#L155)
```fortran
do no = 1,nor...The analysis program ```cov_eq``` computes a subtraction of the "disconnected" part of the correlations [here](https://git.physik.uni-wuerzburg.de/ALF/ALF/-/blob/master/Analysis/cov_eq.F90#L155)
```fortran
do no = 1,norb
do no1 = 1,Norb
bins(m,nb-n_skip)%el(no,no1) = bins(m,nb-n_skip)%el(no,no1) - &
& cmplx(dble(Latt%N),0.d0,kind(0.d0))*Bins0(nb-n_skip,no)*Bins0(nb-n_skip,no1) &
& /Phase(nb-n_skip)
enddo
enddo
```
This subtraction is done on each bin read on the input file, then later the jackknife bins (i.e., everything except 1 bin of MC data) are computed from ```bins```, and the error is computed as usual.
This is not correct. One should do:
* Compute the jaccknife bins (i.e., average on all data except 1 bin) of ```bins``` and of ```Bins0```, *separately*
* Do the subtraction on each jackknife bin
* Compute the standard jackknife formulahttps://git.physik.uni-wuerzburg.de/ALF/ALF/-/issues/138Future improvements to QDR decomposition -> Get SVD branch of stabilization w...2020-06-29T12:45:16ZFlorian GothFuture improvements to QDR decomposition -> Get SVD branch of stabilization working with Jacobi methodWhile reading some papers I came across some developments/explanations for the singular value decompositions.
And in case we ever find ourselves in a position that we need to improve some parts, here are some ideas:
The most natura for u...While reading some papers I came across some developments/explanations for the singular value decompositions.
And in case we ever find ourselves in a position that we need to improve some parts, here are some ideas:
The most natura for us I think is this, interpreting the QR decomposition as preconditioner:
https://doi.org/10.1145/3061709
The paper explaining which SVD Method is superior for resolving small eigenvalues:
https://epubs.siam.org/doi/10.1137/0613074
And a paper reviewing SVDs:
https://www.research.manchester.ac.uk/portal/files/82231456/17m1117732.pdf
Note that there already is zgejsv in lapack which implements a Jacobi SVD with pivoting and preconditioning.