triqs / cthyb Goto Github PK

Dear Oliver and Priyanka,
++++Last time I wrote to you I said;
When I wrote to you last november and you suggested that I should put these issues in the "TRIQS issues"
I assumed from this answer that the DMFT was already a part of TRIQS.
It was a surprise for me last week to find that I have to write my own DMFT program based in the languages used in TRIQS.
So it would be nice to have a program example of a simple compound calculated with WIEN2k, I don't know if there is something like that.
++++You answered;
you should know that DMFT is not a black-box method
and then you added
With the solver in hand, the simplest DMFT loop can be written in about 10 lines.
++++++++++++++++
++++++++++++++++
Today I add;
Yes, I understand and I am still not an expert in DMFT,
but the minimun to start it would be great.
++1- How can one use the WIEN2k output
++2- How to call the CTHYB solver

These issues should be in the DOCUMENTATION;
++A quick introduction to the solver
++The hybridization-expansion algorithm in details
++User Manual
++Implementation details

but as I said, these folders are empty,
so it is quite difficult even for a person that knows DMFT to use the TRIQS package

It would be great to have a program to study the NiO Mott insulator,
but also, for this purpose, Ni or Fe would be sufficient.

I run into a "p not finite" problem during a single band calculation at beta=5000. See the code that generates the error and the error message below.

https://gist.github.com/pseth/e03a67b9dda05e7224ef#file-gistfile1-txt

Hello,

Currently one has to provide a nonzero hybridization function for all blocks enumerated in gf_struct. The only way to include some non-hybridized orbitals would be to omit the corresponding blocks from gf_struct. But then these orbitals cannot enter h_loc, because the fundamental operator set is constructed out of gf_struct.
I think this situation is an artifact of the implementation which should be resolved.

Hi,
using the newest commits of triqs and cthyb gives the error below during compilation.
I'd guess that the error is caused by the triqs commit 572ad4d (For commit f583999 the compilation of cthyb works.).

Regards,
Manuel

[ 8%] Building CXX object c++/CMakeFiles/cthyb_c.dir/atom_diag.cpp.o
/home/manuel/Tools/cthyb/src/c++/atom_diag.cpp: In function ‘std::string triqs::hilbert_space::get_triqs_hdf5_data_scheme(const triqs::hilbert_space::sub_hilbert_space&)’:
/home/manuel/Tools/cthyb/src/c++/atom_diag.cpp:193:14: error: redefinition of ‘std::string triqs::hilbert_space::get_triqs_hdf5_data_scheme(const triqs::hilbert_space::sub_hilbert_space&)’
std::string get_triqs_hdf5_data_scheme(sub_hilbert_space const&) { return "sub_hilbert_space"; }
^
In file included from /home/manuel/Tools/TRIQS/include/triqs/hilbert_space/state.hpp:28:0,
from /home/manuel/Tools/cthyb/src/c++/././config.hpp:28,
from /home/manuel/Tools/cthyb/src/c++/./atom_diag.hpp:22,
from /home/manuel/Tools/cthyb/src/c++/atom_diag.cpp:21:
/home/manuel/Tools/TRIQS/include/triqs/hilbert_space/hilbert_space.hpp:148:21: note: ‘std::string triqs::hilbert_space::get_triqs_hdf5_data_scheme(const triqs::hilbert_space::sub_hilbert_space&)’ previously defined here
friend std::string get_triqs_hdf5_data_scheme(sub_hilbert_space const&) { return "sub_hilbert_space"; }
^
/home/manuel/Tools/cthyb/src/c++/atom_diag.cpp: In function ‘void triqs::hilbert_space::h5_write(triqs::h5::group, const string&, const triqs::hilbert_space::sub_hilbert_space&)’:
/home/manuel/Tools/cthyb/src/c++/atom_diag.cpp:197:7: error: redefinition of ‘void triqs::hilbert_space::h5_write(triqs::h5::group, const string&, const triqs::hilbert_space::sub_hilbert_space&)’
void h5_write(h5::group fg, std::string const& name, sub_hilbert_space const& x) {
^
In file included from /home/manuel/Tools/TRIQS/include/triqs/hilbert_space/state.hpp:28:0,
from /home/manuel/Tools/cthyb/src/c++/././config.hpp:28,
from /home/manuel/Tools/cthyb/src/c++/./atom_diag.hpp:22,
from /home/manuel/Tools/cthyb/src/c++/atom_diag.cpp:21:
/home/manuel/Tools/TRIQS/include/triqs/hilbert_space/hilbert_space.hpp:150:14: note: ‘void triqs::hilbert_space::h5_write(triqs::h5::group, const string&, const triqs::hilbert_space::sub_hilbert_space&)’ previously defined here
friend void h5_write(h5::group fg, std::string const& name, sub_hilbert_space const& hs) {
^
/home/manuel/Tools/cthyb/src/c++/atom_diag.cpp: In function ‘void triqs::hilbert_space::h5_read(triqs::h5::group, const string&, triqs::hilbert_space::sub_hilbert_space&)’:
/home/manuel/Tools/cthyb/src/c++/atom_diag.cpp:204:7: error: redefinition of ‘void triqs::hilbert_space::h5_read(triqs::h5::group, const string&, triqs::hilbert_space::sub_hilbert_space&)’
void h5_read(h5::group fg, std::string const& name, sub_hilbert_space& x) {
^
In file included from /home/manuel/Tools/TRIQS/include/triqs/hilbert_space/state.hpp:28:0,
from /home/manuel/Tools/cthyb/src/c++/././config.hpp:28,
from /home/manuel/Tools/cthyb/src/c++/./atom_diag.hpp:22,
from /home/manuel/Tools/cthyb/src/c++/atom_diag.cpp:21:
/home/manuel/Tools/TRIQS/include/triqs/hilbert_space/hilbert_space.hpp:156:14: note: ‘void triqs::hilbert_space::h5_read(triqs::h5::group, const string&, triqs::hilbert_space::sub_hilbert_space&)’ previously defined here
friend void h5_read(h5::group fg, std::string const& name, sub_hilbert_space & hs) {
^
make[2]: *** [c++/CMakeFiles/cthyb_c.dir/atom_diag.cpp.o] Error 1
make[1]: *** [c++/CMakeFiles/cthyb_c.dir/all] Error 2
make: *** [all] Error 2

On my mac two tests fail: kanamori_kwargs and kanamori_offdiag.
However, the differences seem to be small. For kanamori_kwargs the test passes when I set the precision to "-d 1e-6", for kanamori_offdiag when I set it to "-d 1e-5" (for 1e-6 it still fails).
On debian stable the tests pass with 1e-7 (same commit).

while performing an install cmake complains

$cmake -DTRIQS_PATH=/home/oscar/libs/ ../cthyb/
-- The CXX compiler identification is GNU 4.9.2
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- TRIQS : Adding compilation flags detected by the library (C++11/14, libc++, etc...) 
-- Found TRIQS: /home/oscar/libs/lib/libtriqs.so  
-- Hash: d275c6453168e90ddbf7b88cef977b4e35fab5be
-- Preparing extension module cthyb
-- Configuring done
CMake Warning (dev) at /home/oscar/libs/share/triqs/cmake/TRIQSConfig.cmake:111 (add_dependencies):
  Policy CMP0046 is not set: Error on non-existent dependency in
  add_dependencies.  Run "cmake --help-policy CMP0046" for policy details.
  Use the cmake_policy command to set the policy and suppress this warning.

  The dependency target "/home/oscar/dev/buicthyb/python/cthyb_desc.py" of
  target "python_wrap_cthyb" does not exist.
Call Stack (most recent call first):
  python/CMakeLists.txt:12 (triqs_python_extension)
This warning is for project developers.  Use -Wno-dev to suppress it.

-- Generating done
-- Build files have been written to: /home/oscar/dev/buicthyb

but it finishes. Nevertheless make is not able to find the include dir either

$make
Scanning dependencies of target cthyb_c
[  4%] Building CXX object c++/CMakeFiles/cthyb_c.dir/solver_core.cpp.o
In file included from /home/oscar/dev/cthyb/c++/solver_core.cpp:21:0:
/home/oscar/dev/cthyb/c++/./solver_core.hpp:22:30: fatal error: triqs/mc_tools.hpp: No such file or directory
 #include <triqs/mc_tools.hpp>
                              ^
compilation terminated.
c++/CMakeFiles/cthyb_c.dir/build.make:54: recipe for target 'c++/CMakeFiles/cthyb_c.dir/solver_core.cpp.o' failed
make[2]: *** [c++/CMakeFiles/cthyb_c.dir/solver_core.cpp.o] Error 1
CMakeFiles/Makefile2:75: recipe for target 'c++/CMakeFiles/cthyb_c.dir/all' failed
make[1]: *** [c++/CMakeFiles/cthyb_c.dir/all] Error 2
Makefile:126: recipe for target 'all' failed
make: *** [all] Error 2

After months of struggling with the installation of the TRIQS I finally got it running.
I observed that there are a few omissions in the example.py file:
"import matplotlib
matplotlib.pyplot.show()"
It was only after appending the above two lines of code that example.py produced a graph.

In the LDA+DMFT Wien2k was used as model which would have been alright for everybody if Wien2k was a free software, but as it is not free. Hence TRIQS would fill a really big void for condensed matter physics community if a freely available software like QUANTUM ESPRESSO can be used to show how DFT is incorporated into TRIQS.
I know it is a lot of sacrifice but there is need for it.

Hi
I got folowing error during make:::

/home/soumen/Multi_orbital/cthyb-master/c++/./atomic_correlators_worker.checks.cpp:54:21: error: ‘make_unit_matrix’ was not declared in this scope
matrix M = make_unit_matrix(get_block_dim(b));
^
/home/soumen/Multi_orbital/cthyb-master/c++/./atomic_correlators_worker.checks.cpp:54:38: error: expected primary-expression before ‘double’
matrix M = make_unit_matrix(get_block_dim(b));
^
make[2]: *** [c++/CMakeFiles/cthyb_c.dir/atomic_correlators_worker.cpp.o] Error 1
make[1]: *** [c++/CMakeFiles/cthyb_c.dir/all] Error 2
make: *** [all] Error 2

I was searching around the documentation and I could not find any hint on how to measure the double occupation during the Simulation.
How can this information acquired?

This test is broken. Need to figure out when (which patch) and why.

Hi,
I am trying to compile cthyb on my Macbook (OSX 10.11.3) using Apple's LLVM plus Macport python2.7. I am trying to compile in C++11/14 mode in order to build the documentation.

I did compile and deploy Triqs successfully, all tests were passed. The make configure of cthyb passes with a (unrelated?) warning message:
Syntax Warning in cmake code at
/Users/gabriele/Software/Triqs/Install/1.4dev/share/triqs/cmake/TRIQSConfig.cmake:41:168
Argument not separated from preceding token by whitespace.

However, when I try to make cthyb, I got this error:
...
[ 19%] Generating cthyb_desc.py, cthyb_wrap.cpp, ../include/pytriqs/converters/cthyb.hpp
Traceback (most recent call last):
File "/Users/gabriele/Software/Triqs/Builds/cthyb-1.4dev/python/cthyb_desc.py", line 9, in
module.use_module('gf', 'triqs')
TypeError: use_module() takes exactly 2 arguments (3 given)
make[2]: *** [python/cthyb_desc.py] Error 1
make[2]: *** Deleting file `python/cthyb_desc.py'
make[1]: *** [python/CMakeFiles/cthyb.dir/all] Error 2
make: *** [all] Error 2

Any idea?

I got error messages when I tried to compile CTHYB in the master branch with options,
Complex interaction-ON, Complex hybridization-ON.

When I tried to compile CTHYB without complex options, there was no problem.

Please check following information.

• git hash for cthyb
  3c18353
• git has for triqs
  416eb258dbdc39b5736e8633da5289aafe3a8147
• error message

-- Hash: 3c18353
-- TRIQS : Adding compilation flags detected by the library (C++11/14, libc++, etc...)
-- Preparing extension module cthyb
-- Configuring done
-- Generating done
-- Build files have been written to: /home/minjae/METHODS/TRIQS/cthyb/build
Scanning dependencies of target py_copy
[ 4%] Generating py_copy.tar
[ 4%] Built target py_copy
Scanning dependencies of target cthyb_c
[ 17%] [ 17%] [ 17%] Building CXX object c++/CMakeFiles/cthyb_c.dir/atom_diag.cpp.o
[ 26%] [ 26%] Building CXX object c++/CMakeFiles/cthyb_c.dir/atom_diag_functions.cpp.o
Building CXX object c++/CMakeFiles/cthyb_c.dir/solver_core.cpp.o
[ 30%] Building CXX object c++/CMakeFiles/cthyb_c.dir/impurity_trace.cpp.o
Building CXX object c++/CMakeFiles/cthyb_c.dir/atom_diag_worker.cpp.o
Building CXX object c++/CMakeFiles/cthyb_c.dir/measure_density_matrix.cpp.o
Linking CXX static library libcthyb_c.a
[ 30%] Built target cthyb_c
Scanning dependencies of target h_diag_test
Scanning dependencies of target anderson_block_qn
Scanning dependencies of target kanamori
Scanning dependencies of target anderson
Scanning dependencies of target anderson_qn
Scanning dependencies of target anderson_block
Scanning dependencies of target atomic_gf
Scanning dependencies of target cthyb
[ 34%] Building CXX object test/c++/CMakeFiles/h_diag_test.dir/h_diag_test.cpp.o
[ 60%] [ 47%] [ 60%] [ 60%] [ 60%] [ 60%] [ 65%] Building CXX object test/c++/CMakeFiles/anderson_block_qn.dir/anderson.cpp.o
Building CXX object test/c++/CMakeFiles/kanamori.dir/kanamori.cpp.o
Building CXX object test/c++/CMakeFiles/anderson.dir/anderson.cpp.o
Building CXX object test/c++/CMakeFiles/anderson_qn.dir/anderson.cpp.o
Building CXX object test/c++/CMakeFiles/atomic_gf.dir/atomic_gf.cpp.o
Building CXX object test/c++/CMakeFiles/anderson_block.dir/anderson.cpp.o
Building CXX object python/CMakeFiles/cthyb.dir/cthyb/cthyb_wrap.cpp.o
/home/minjae/METHODS/TRIQS/cthyb/src/test/c++/h_diag_test.cpp: In function ‘int main()’:
/home/minjae/METHODS/TRIQS/cthyb/src/test/c++/h_diag_test.cpp:34:37: error: no matching function for call to ‘cthyb::atom_diag::atom_diag(triqs::operators::many_body_operator_generictriqs::utility::real_or_complex&, triqs::hilbert_space::fundamental_operator_set&, std::vectortriqs::operators::many_body_operator_generic&)’
atom_diag h_diag(H, fops, qn_list);
^
/home/minjae/METHODS/TRIQS/cthyb/src/test/c++/h_diag_test.cpp:34:37: note: candidates are:
In file included from /home/minjae/METHODS/TRIQS/cthyb/src/test/c++/h_diag_test.cpp:2:0:
/home/minjae/METHODS/TRIQS/cthyb/src/c++/./atom_diag.hpp:45:2: note: cthyb::atom_diag::atom_diag(const many_body_op_t&, const triqs::hilbert_space::fundamental_operator_set&, const std::vectortriqs::operators::many_body_operator_generic<std::complex >&)
atom_diag(many_body_op_t const& h_, fundamental_operator_set const& fops, std::vector<many_body_op_t> const& qn_vector);
^
/home/minjae/METHODS/TRIQS/cthyb/src/c++/./atom_diag.hpp:45:2: note: no known conversion for argument 3 from ‘std::vectortriqs::operators::many_body_operator_generic’ to ‘const std::vectortriqs::operators::many_body_operator_generic<std::complex >&’
/home/minjae/METHODS/TRIQS/cthyb/src/c++/./atom_diag.hpp:44:2: note: cthyb::atom_diag::atom_diag(const many_body_op_t&, const triqs::hilbert_space::fundamental_operator_set&)
atom_diag(many_body_op_t const& h_, fundamental_operator_set const& fops);
^
/home/minjae/METHODS/TRIQS/cthyb/src/c++/./atom_diag.hpp:44:2: note: candidate expects 2 arguments, 3 provided
/home/minjae/METHODS/TRIQS/cthyb/src/c++/./atom_diag.hpp:43:22: note: cthyb::atom_diag::atom_diag()
TRIQS_CPP2PY_IGNORE atom_diag() = default;
^
/home/minjae/METHODS/TRIQS/cthyb/src/c++/./atom_diag.hpp:43:22: note: candidate expects 0 arguments, 3 provided
/home/minjae/METHODS/TRIQS/cthyb/src/c++/./atom_diag.hpp:34:7: note: cthyb::atom_diag::atom_diag(const cthyb::atom_diag&)
class atom_diag {
^
/home/minjae/METHODS/TRIQS/cthyb/src/c++/./atom_diag.hpp:34:7: note: candidate expects 1 argument, 3 provided
/home/minjae/METHODS/TRIQS/cthyb/src/c++/./atom_diag.hpp:34:7: note: cthyb::atom_diag::atom_diag(cthyb::atom_diag&&)
/home/minjae/METHODS/TRIQS/cthyb/src/c++/./atom_diag.hpp:34:7: note: candidate expects 1 argument, 3 provided
make[2]: *** [test/c++/CMakeFiles/h_diag_test.dir/h_diag_test.cpp.o] Error 1
make[1]: *** [test/c++/CMakeFiles/h_diag_test.dir/all] Error 2
make[1]: *** Waiting for unfinished jobs....
Linking CXX executable anderson_block
[ 65%] Built target anderson_block
Linking CXX executable anderson
Linking CXX executable anderson_qn
Linking CXX executable anderson_block_qn
[ 65%] Built target anderson
Linking CXX executable atomic_gf
Linking CXX shared module cthyb/cthyb.so
[ 65%] Built target anderson_block_qn
[ 65%] Built target anderson_qn
Linking CXX executable kanamori
[ 69%] Built target cthyb
[ 69%] Built target atomic_gf
[ 69%] Built target kanamori
make: *** [all] Error 2

Using the complex branch, hash de4ab25, the solver compiles (with gcc 4.9.3) when both HYBRIDISATION_IS_COMPLEX and LOCAL_HAMILTONIAN_IS_COMPLEX are ON.
However, if I only turn on one of them, I get errors during the compilation.

When turning on only HYBRIDISATION_IS_COMPLEX:

Building CXX object c++/CMakeFiles/cthyb_c.dir/measure_density_matrix.cpp.o

[...]

In file included from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/arrays/impl/indexmap_storage_pair.hpp:24:0,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/arrays/array.hpp:25,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/arrays.hpp:27,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/././././config.hpp:22,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/./././atom_diag.hpp:22,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/./././configuration.hpp:22,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/././impurity_trace.hpp:22,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/././qmc_data.hpp:22,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/./measure_density_matrix.hpp:22,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/measure_density_matrix.cpp:21:
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/arrays/impl/./assignment.hpp: In instantiation of ‘static void triqs::arrays::assignment::_ops_<A, B, 'A'>::invoke(A&, const B&) [with A = double; B = std::complex<double>]’:
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/arrays/impl/./assignment.hpp:121:163:   required from ‘void triqs::arrays::assignment::impl<LHS, RHS, OP, typename boost::enable_if_c<(((triqs::arrays::ImmutableCuboidArray<RHS>::value && (! triqs::arrays::is_scalar_for<RHS, LHS>::value)) && (! triqs::arrays::assignment::is_isp<RHS, LHS>::value)) && (! triqs::arrays::assignment::is_special<RHS, LHS>::value)), void>::type>::operator()(const Args& ...) const [with Args = {long int, long int}; LHS = triqs::arrays::matrix<double>; RHS = triqs::arrays::matrix_expr<triqs::utility::tags::multiplies, triqs::arrays::_scalar_wrap<std::complex<double>, false>, const triqs::arrays::matrix<double>&>; char OP = 'A'; typename boost::enable_if_c<(((triqs::arrays::ImmutableCuboidArray<RHS>::value && (! triqs::arrays::is_scalar_for<RHS, LHS>::value)) && (! triqs::arrays::assignment::is_isp<RHS, LHS>::value)) && (! triqs::arrays::assignment::is_special<RHS, LHS>::value)), void>::type = void]’
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/arrays/impl/./../indexmaps/cuboid/foreach.hpp:77:4:   required from ‘void triqs::arrays::indexmaps::cuboid::foreach_impl(triqs::arrays::_traversal_c, const triqs::arrays::indexmaps::cuboid::domain_t<2>&, const triqs::arrays::memory_layout<2>&, FntType) [with FntType = triqs::arrays::assignment::impl<triqs::arrays::matrix<double>, triqs::arrays::matrix_expr<triqs::utility::tags::multiplies, triqs::arrays::_scalar_wrap<std::complex<double>, false>, const triqs::arrays::matrix<double>&>, 'A', void>]’
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/arrays/impl/./../indexmaps/cuboid/foreach.hpp:107:95:   required from ‘std::c14::enable_if_t<triqs::arrays::ImmutableCuboidArray<T>::value> triqs::arrays::foreach(const T&, const Function&) [with T = triqs::arrays::matrix<double>; Function = triqs::arrays::assignment::impl<triqs::arrays::matrix<double>, triqs::arrays::matrix_expr<triqs::utility::tags::multiplies, triqs::arrays::_scalar_wrap<std::complex<double>, false>, const triqs::arrays::matrix<double>&>, 'A', void>; std::c14::enable_if_t<triqs::arrays::ImmutableCuboidArray<T>::value> = void]’
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/arrays/impl/./assignment.hpp:122:52:   required from ‘void triqs::arrays::assignment::impl<LHS, RHS, OP, typename boost::enable_if_c<(((triqs::arrays::ImmutableCuboidArray<RHS>::value && (! triqs::arrays::is_scalar_for<RHS, LHS>::value)) && (! triqs::arrays::assignment::is_isp<RHS, LHS>::value)) && (! triqs::arrays::assignment::is_special<RHS, LHS>::value)), void>::type>::invoke() [with LHS = triqs::arrays::matrix<double>; RHS = triqs::arrays::matrix_expr<triqs::utility::tags::multiplies, triqs::arrays::_scalar_wrap<std::complex<double>, false>, const triqs::arrays::matrix<double>&>; char OP = 'A'; typename boost::enable_if_c<(((triqs::arrays::ImmutableCuboidArray<RHS>::value && (! triqs::arrays::is_scalar_for<RHS, LHS>::value)) && (! triqs::arrays::assignment::is_isp<RHS, LHS>::value)) && (! triqs::arrays::assignment::is_special<RHS, LHS>::value)), void>::type = void]’
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/arrays/impl/./assignment.hpp:44:4:   required from ‘void triqs::arrays::triqs_arrays_compound_assign_delegation(LHS&, const RHS&, char_<OP>) [with LHS = triqs::arrays::matrix<double>; RHS = triqs::arrays::matrix_expr<triqs::utility::tags::multiplies, triqs::arrays::_scalar_wrap<std::complex<double>, false>, const triqs::arrays::matrix<double>&>; char OP = 'A'; char_<OP> = std::integral_constant<char, 'A'>]’
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/arrays/matrix.hpp:237:3:   required from ‘triqs::arrays::matrix<ValueType, TraversalOrder>& triqs::arrays::matrix<ValueType, TraversalOrder>::operator+=(const RHS&) [with RHS = triqs::arrays::matrix_expr<triqs::utility::tags::multiplies, triqs::arrays::_scalar_wrap<std::complex<double>, false>, const triqs::arrays::matrix<double>&>; ValueType = double; TraversalOrder = void]’
/home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/measure_density_matrix.cpp:52:16:   required from here
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/arrays/impl/./assignment.hpp:65:103: error: no match for ‘operator+=’ (operand types are ‘double’ and ‘const std::complex<double>’)
   template<typename A,typename B> struct _ops_ <A,B,'A'> { static void invoke (A & a, B const & b) { a+=b;} };

When turning on only LOCAL_HAMILTONIAN_IS_COMPLEX:

Building CXX object c++/CMakeFiles/cthyb_c.dir/solver_core.cpp.o

[...]
In file included from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/solver_core.cpp:33:0:
/home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/measure_g.hpp: In lambda function:
/home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/measure_g.hpp:56:80: error: no match for ‘operator+=’ (operand types are ‘double’ and ‘std::complex<double>’)
    this->g_tau[closest_mesh_pt(double(y.first - x.first))](y.second, x.second) +=
                                                                                ^
/home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/measure_g.hpp:56:80: note: candidates are:
In file included from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/defs.hpp:30:0,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/gf_classes.hpp:23,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/gfs.hpp:26,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/././config.hpp:23,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/./solve_parameters.hpp:2,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/./solver_core.hpp:25,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/solver_core.cpp:21:
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/../singularity/tail.hpp:344:29: note: template<class T> void triqs::gfs::operator+=(triqs::gfs::tail_view, T&&)
  template <typename T> void operator+=(tail_view g, T &&x) { g = g + x; }
                             ^
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/../singularity/tail.hpp:344:29: note:   template argument deduction/substitution failed:
In file included from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/solver_core.cpp:33:0:
/home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/measure_g.hpp:56:80: note:   cannot convert ‘this->cthyb::measure_g::g_tau.triqs::gfs::gf_view<Mesh, Target, Singularity, Evaluator>::operator[]<{double}>((*(const triqs::gfs::closest_pt_wrap<double>*)(& triqs::gfs::closest_mesh_pt<{double}>((* & triqs::utility::operator-((* & y.std::pair<triqs::utility::time_pt, int>::first), (* & x.std::pair<triqs::utility::time_pt, int>::first)).triqs::utility::time_pt::operator double()))))).triqs::arrays::matrix_tensor_proxy<A, IsMatrix>::operator()<{const int&, const int&}>((* & y.std::pair<triqs::utility::time_pt, int>::second), (* & x.std::pair<triqs::utility::time_pt, int>::second))’ (type ‘double’) to type ‘triqs::gfs::tail_view’
    this->g_tau[closest_mesh_pt(double(y.first - x.first))](y.second, x.second) +=
                                                                                ^
In file included from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/defs.hpp:30:0,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/gf_classes.hpp:23,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/gfs.hpp:26,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/././config.hpp:23,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/./solve_parameters.hpp:2,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/./solver_core.hpp:25,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/solver_core.cpp:21:
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/../singularity/tail.hpp:349:29: note: template<class T> void triqs::gfs::operator+=(triqs::gfs::tail&, T&&)
  template <typename T> void operator+=(tail &g, T &&x) { g = g + x; }
                             ^
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/../singularity/tail.hpp:349:29: note:   template argument deduction/substitution failed:
In file included from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/solver_core.cpp:33:0:
/home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/measure_g.hpp:56:80: note:   cannot convert ‘this->cthyb::measure_g::g_tau.triqs::gfs::gf_view<Mesh, Target, Singularity, Evaluator>::operator[]<{double}>((*(const triqs::gfs::closest_pt_wrap<double>*)(& triqs::gfs::closest_mesh_pt<{double}>((* & triqs::utility::operator-((* & y.std::pair<triqs::utility::time_pt, int>::first), (* & x.std::pair<triqs::utility::time_pt, int>::first)).triqs::utility::time_pt::operator double()))))).triqs::arrays::matrix_tensor_proxy<A, IsMatrix>::operator()<{const int&, const int&}>((* & y.std::pair<triqs::utility::time_pt, int>::second), (* & x.std::pair<triqs::utility::time_pt, int>::second))’ (type ‘double’) to type ‘triqs::gfs::tail&’
    this->g_tau[closest_mesh_pt(double(y.first - x.first))](y.second, x.second) +=
                                                                                ^
In file included from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/gf_classes.hpp:28:0,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/gfs.hpp:26,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/././config.hpp:23,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/./solve_parameters.hpp:2,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/./solver_core.hpp:25,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/solver_core.cpp:21:
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/gf_expr.hpp:160:7: note: template<class Mesh, class Target, class Evaluator, class T> void triqs::gfs::operator+=(triqs::gfs::gf_view<Mesh, Target, Evaluator>, const T&)
  void operator OP1(gf_view<Mesh, Target, Evaluator> g, T const &x) {                                                         \
       ^
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/gf_expr.hpp:168:2: note: in expansion of macro ‘DEFINE_OPERATOR’
  DEFINE_OPERATOR(+=, +);
  ^
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/gf_expr.hpp:160:7: note:   template argument deduction/substitution failed:
  void operator OP1(gf_view<Mesh, Target, Evaluator> g, T const &x) {                                                         \
       ^
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/gf_expr.hpp:168:2: note: in expansion of macro ‘DEFINE_OPERATOR’
  DEFINE_OPERATOR(+=, +);
  ^
In file included from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/solver_core.cpp:33:0:
/home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/measure_g.hpp:56:80: note:   mismatched types ‘triqs::gfs::gf_view<Mesh, Target, Evaluator>’ and ‘double’
    this->g_tau[closest_mesh_pt(double(y.first - x.first))](y.second, x.second) +=
                                                                                ^
In file included from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/gf_classes.hpp:28:0,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/gfs.hpp:26,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/././config.hpp:23,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/./solve_parameters.hpp:2,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/./solver_core.hpp:25,
                 from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/solver_core.cpp:21:
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/gf_expr.hpp:164:7: note: template<class Mesh, class Target, class Evaluator, class T> void triqs::gfs::operator+=(triqs::gfs::gf<M, T, S>&, const T&)
  void operator OP1(gf<Mesh, Target, Evaluator> &g, T const &x) {                                                             \
       ^
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/gf_expr.hpp:168:2: note: in expansion of macro ‘DEFINE_OPERATOR’
  DEFINE_OPERATOR(+=, +);
  ^
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/gf_expr.hpp:164:7: note:   template argument deduction/substitution failed:
  void operator OP1(gf<Mesh, Target, Evaluator> &g, T const &x) {                                                             \
       ^
/home/lv70759/gkraberger/progs/TRIQS_complex/install/include/triqs/./gfs/./impl/gf_expr.hpp:168:2: note: in expansion of macro ‘DEFINE_OPERATOR’
  DEFINE_OPERATOR(+=, +);
  ^
In file included from /home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/solver_core.cpp:33:0:
/home/lv70759/gkraberger/progs/TRIQS_complex/cthyb.src/c++/measure_g.hpp:56:80: note:   mismatched types ‘triqs::gfs::gf<M, T, S>’ and ‘double’
    this->g_tau[closest_mesh_pt(double(y.first - x.first))](y.second, x.second) +=
                                                                                ^

This is just the first error in make, so there might be more errors later on.

In my code for dft_tools I use:
S.G0_iw <<= inverse(S.Sigma_iw + inverse(S.G_iw))
and
S.G0_iw = mpi.bcast(S.G0_iw)
but it is not possible:

AttributeError: attribute 'G0_iw' of 'pytriqs.applications.impurity_solvers.cthyb.SolverCore' objects is not writable

Why?

Using
LLVM version 6.0 (clang-600.0.51) (based on LLVM 3.5svn)
the compilation of hilbert_test fails with

[ 56%] Building CXX object test/c++/CMakeFiles/hilbert_test.dir/hilbert_test.cpp.o
/Users/aichhorn/work/Codes/TRIQS1.x/CTHYB/test/c++/hilbert_test.cpp:44:66: error: chosen constructor is explicit in
copy-initialization
std::cout << "fock state for vacuum = " << h.get_fock_state(f1,{}) << std::endl;
^~
/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/../include/c++/v1/set:428:14: note:
constructor declared here
explicit set(const value_compare& __comp = value_compare())
^
/Users/aichhorn/work/Codes/TRIQS1.x/CTHYB/test/c++/../../c++/triqs/draft/hilbert_space_tools/hilbert_space.hpp:60:121: note:
passing argument to parameter 'indices' here
...get_fock_state(fundamental_operator_set const &fops, std::set<fundamental_operator_set::indices_t> const& indices) const {
^
1 error generated.
make[2]: *** [test/c++/CMakeFiles/hilbert_test.dir/hilbert_test.cpp.o] Error 1
make[1]: *** [test/c++/CMakeFiles/hilbert_test.dir/all] Error 2
make: *** [all] Error 2

Does not occur on the linux machine, though.

I installed the newest version of the TRIQS library and the newest cthyb solver. Installation works,however the anderson test (nr 3) starts, but terminates after 600 sec.

This is the output in anderson_output:
Welcome to the CTHYB solver
The local Hamiltonian of the problem:
-1_C^+(tot,0)C(tot,0) + -1_C^+(tot,1)C(tot,1) + 2*C^+(tot,0)C^+(tot,1)C(tot,1)C(tot,0)
Using autopartition algorithm to partition the local Hilbert space
Found 4 subspaces.

1%; 2%;

anderson_output.err is empty.
And no it is not my computer - I tried it on another machine. ;)

Reminder to self.

Hello,

In commit 58c4bcd I added a new (advanced) cmake option TRIQS_INVENTORY_DIR set to ${TRIQS_PATH}/share/doc/triqs by default.

This option tells Sphinx where it should look for an inventory file needed to generate external references to the library docs. According to the default value, these references will point at the locally installed TRIQS docs.
However, one has to give an explicit URL, if the docs are built for a web-site:

-DTRIQS_INVENTORY_DIR=http://ipht.cea.fr/triqs/1.2/

implying http://ipht.cea.fr/triqs/1.2/objects.inv is the correct Sphinx inventory file.

There is an optimization in impurity_trace, which makes the algorithm skip updating the tree root in some (unknown to me) cases.

Yes, I know that everyone is aware of this bug.
I'm creating this issue in a desperate attempt to draw attention to it. As long as the bug is not fixed, I cannot proceed with measure_static, and my current physical project remains completely blocked.

The code compiles fine, but when runnning the tests, I get the following error messages:

terminate called after throwing an instance of 'boost::exception_detail::clone_implboost::exception_detail::error_info_injector<boost::mpi::exception >'
what(): MPI_Allreduce: Invalid buffer pointer, error stack:
MPI_Allreduce(855): MPI_Allreduce(sbuf=0x2244c28, rbuf=0x2244c28, count=1, MPI_LONG, op=0x98000000, MPI_COMM_WORLD) failed
MPI_Allreduce(830): Buffers must not be aliased

Up to this point, the code runs fine, the output file looks normal.

I am using debian stable, gcc 4.8.2, boost 1.54. The same error occured with boost 1.49.

Hi,

On our cluster, many tests die with segfaults using cthyb 53c80ee. Using commit bf8e132 everything is fine. TRIQS version is TRIQS/triqs@b441431.
Strangely, this problem does not happen on our local machines.

The following tests FAILED:
          3 - anderson (SEGFAULT)
          4 - anderson_block (SEGFAULT)
          5 - anderson_qn (SEGFAULT)
          6 - anderson_block_qn (SEGFAULT)
          7 - spinless (SEGFAULT)
          8 - spinless_qn (SEGFAULT)
          9 - kanamori (SEGFAULT)
         10 - kanamori_qn (SEGFAULT)
         11 - kanamori_offdiag (SEGFAULT)
         12 - kanamori_offdiag_qn (SEGFAULT)
         13 - legendre (SEGFAULT)
         14 - atomic_gf (SEGFAULT)
         16 - atomic_observables (Failed)
         17 - kanamori (Failed)
         18 - slater (Failed)
         19 - measure_static (Failed)
         20 - histograms (Failed)
         21 - move_global (Failed)

The error message when running, e.g. anderson, is

build/cthyb/test/c++ $ ./anderson
[==========] Running 1 test from 1 test case.
[----------] Global test environment set-up.
[----------] 1 test from CtHyb
[ RUN      ] CtHyb.Anderson
Welcome to the CTHYB solver
The local Hamiltonian of the problem:
(-1,0)*C^+(tot,0)C(tot,0) + (-1,0)*C^+(tot,1)C(tot,1) + (2,0)*C^+(tot,0)C^+(tot,1)C(tot,1)C(tot,0)
Using autopartition algorithm to partition the local Hilbert space
Found 4 subspaces.

Warming up ...
2%; 4%; 6%; 8%; 10%; 12%; 14%; 16%; 18%; 20%; 22%; 24%; 26%; 28%; 30%; 32%; 34%; 36%; 38%; 40%; 42%; 44%; 46%; 48%; 51%; 53%; 55%; 57%; 59%; 61%; 63%; 65%; 67%; 69%; 71%; 73%; 75%; 77%; 79%; 81%; 83%; 85%; 87%; 89%; 91%; 93%; 95%; 97%; 100%; 
Accumulating ...
1%; 2%; 3%; 4%; 5%; 6%; 7%; 8%; 9%; 10%; 11%; 12%; 13%; 14%; 15%; 16%; 17%; 18%; 19%; 20%; 21%; 22%; 23%; 24%; 25%; 26%; 27%; 28%; 29%; 30%; 31%; 32%; 33%; 34%; 35%; 36%; 37%; 38%; 39%; 40%; 41%; 42%; 43%; 44%; 45%; 46%; 47%; 48%; 49%; 50%; 51%; 52%; 53%; 54%; 55%; 56%; 57%; 58%; 59%; 60%; 61%; 62%; 63%; 64%; 65%; 66%; 67%; 68%; 69%; 70%; 71%; 72%; 73%; 74%; 75%; 76%; 77%; 78%; 79%; 80%; 81%; 82%; 83%; 84%; 85%; 86%; 87%; 88%; 89%; 90%; 91%; 92%; 93%; 94%; 95%; 96%; 97%; 98%; 99%; 100%; 

[Node 0] Acceptance rate for all moves:
Move set Insert two operators: 0.043997
  Move  Insert Delta_tot: 0.043997
Move set Remove two operators: 0.0438907
  Move  Remove Delta_tot: 0.0438907
Move  Shift one operator: 0.242194
[Node 0] Simulation lasted: 0 seconds
[Node 0] Number of measures: 5000
Total number of measures: 5000
Average sign: (1,0)
[l31:22254] *** Process received signal ***
[l31:22254] Signal: Segmentation fault (11)
[l31:22254] Signal code: Address not mapped (1)
[l31:22254] Failing at address: 0xfffff9a0e9000018
[l31:22254] [ 0] /lib64/libpthread.so.0(+0xf710)[0x2b10f4f69710]
[l31:22254] [ 1] ./anderson(_ZNSt8_Rb_treeISsSt4pairIKSsN5triqs10statistics9histogramEESt10_Select1stIS5_ESt4lessISsESaIS5_EE8_M_eraseEPSt13_Rb_tree_nodeIS5_E+0x38)[0x421128]
[l31:22254] [ 2] ./anderson(_ZNSt8_Rb_treeISsSt4pairIKSsN5triqs10statistics9histogramEESt10_Select1stIS5_ESt4lessISsESaIS5_EE8_M_eraseEPSt13_Rb_tree_nodeIS5_E+0x44)[0x421134]
[l31:22254] [ 3] ./anderson(_ZN5cthyb11solver_coreD2Ev+0x1f)[0x4291cf]
[l31:22254] [ 4] ./anderson(_ZN19CtHyb_Anderson_Test8TestBodyEv+0xf39)[0x41ba49]
[l31:22254] [ 5] /home/lv70759/gkraberger/progs/TRIQS_complex/lib/libgtest.so(_ZN7testing8internal35HandleExceptionsInMethodIfSupportedINS_4TestEvEET0_PT_MS4_FS3_vEPKc+0x33)[0x2b10f74198c3]
[l31:22254] [ 6] /home/lv70759/gkraberger/progs/TRIQS_complex/lib/libgtest.so(_ZN7testing4Test3RunEv+0xba)[0x2b10f741066a]
[l31:22254] [ 7] /home/lv70759/gkraberger/progs/TRIQS_complex/lib/libgtest.so(_ZN7testing8TestInfo3RunEv+0x118)[0x2b10f74107b8]
[l31:22254] [ 8] /home/lv70759/gkraberger/progs/TRIQS_complex/lib/libgtest.so(_ZN7testing8TestCase3RunEv+0xb5)[0x2b10f7410895]
[l31:22254] [ 9] /home/lv70759/gkraberger/progs/TRIQS_complex/lib/libgtest.so(_ZN7testing8internal12UnitTestImpl11RunAllTestsEv+0x260)[0x2b10f7410e10]
[l31:22254] [10] /home/lv70759/gkraberger/progs/TRIQS_complex/lib/libgtest.so(_ZN7testing8UnitTest3RunEv+0x4c)[0x2b10f74110fc]
[l31:22254] [11] ./anderson(main+0x50)[0x41f980]
[l31:22254] [12] /lib64/libc.so.6(__libc_start_main+0xfd)[0x2b10f785fd5d]
[l31:22254] [13] ./anderson[0x419f79]
[l31:22254] *** End of error message ***
Segmentation fault

C++ compiler: gcc (GCC) 4.9.3 (on the cluster), c++ (Debian 4.9.2-10) 4.9.2 (local)

I realize that this is probably not enough info to find the root of this problem, but I cannot really tell what else is important. I'm happy to provide more info...

Gernot

cthyb throws an exception when I try to set the partition method to none.

Traceback (most recent call last):
  File "spin_mixing.py", line 60, in <module>
    S.solve(h_int=H, **p)
RuntimeError: .. Error occurred at Tue Feb  9 15:04:41 2016

.. Error .. calling C++ overload 
.. solve() -> void 
.. in implementation of method SolverCore.solve
.. C++ error was : 
.. Triqs runtime error at /home/igor/Physics/TRIQS/cthyb.git/c++/atom_diag_worker.cpp : 361

The bare vacuum is not in a block of size 1 !
.. Error occurred on node 0

Is there a good reason for this restriction? It implies that the total number of particles is a quantum number of h_loc, which is not necessarily the case.

I have pulled the latest version of the solver (d275c64) recompiled it and run tests. Practically all tests failed because of the difference in the output h5-files. For one of the tests (anderson) I checked the content of the output.h5 and the data of Gtot turned out to be just a zero array.

Is there anything extra I can check to identify the problem?

The following tests FAILED:
3 - anderson (Failed)
4 - anderson_block (Failed)
5 - anderson_qn (Failed)
6 - anderson_block_qn (Failed)
7 - spinless (Failed)
8 - spinless_qn (Failed)
9 - kanamori (Failed)
10 - kanamori_qn (Failed)
11 - kanamori_offdiag (Failed)
12 - kanamori_offdiag_qn (Failed)
13 - legendre (Failed)
15 - kanamori (Failed)
16 - single_site_bethe (Failed)
17 - slater (Failed)

When trying to install the compiled code, OS X complains about duplicating the run time path in the libraries:

-- Installing: /Users/aichhorn/work/progs/TRIQS1.2dev/lib/libcthyb_c.dylib
error: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/install_name_tool: for: /Users/aichhorn/work/progs/TRIQS1.2dev/lib/libcthyb_c.dylib (for architecture x86_64) option "-add_rpath /Users/aichhorn/work/progs/TRIQS1.2dev/lib" would duplicate path, file already has LC_RPATH for: /Users/aichhorn/work/progs/TRIQS1.2dev/lib

Removing the according command in c++/cmake_install.cmake (and accordingly in python/cmake_install.cmake) removes this problem. The installed code seems to work.

I am using XCode version 6.0.1

Hi,
I need the energy-spectrum that is used by solve() for some checks in my pytriqs application.
Can you make the spectrum accessible for the python-interface?

Dear all,

we ran into a segmentation fault in cthyb starting from a particular G0 [6x6 block Gf] employing dfttools and the simple hk_converter. If there is a small shift in the k-mesh away from the high-symmetry points, the auto-partition method seems to fail (uninitialized variables) and the subsequently called complete_init creates a segmentation fault:

BT >>>>>>>>>>>>>>>>>

==11059== Process terminating with default action of signal 11 (SIGSEGV)
==11059== Access not within mapped region at address 0x36CC87800
==11059== at 0x1A28B51A: ZNK5triqs13hilbert_space19imperative_operatorINS0_13hilbert_spaceEdLb0EEclINS0_5stateINS0_17sub_hilbert_spaceEdLb0EEEJEEET_RKS8_DpOT0 (in /home/hoeppner/src/TRIQS/install_triqs/lib/libcthyb_c.so)
==11059== by 0x1A2846B2: cthyb::sorted_spaces::complete_init(triqs::utility::many_body_operator const&) (in /home/hoeppner/src/TRIQS/install_triqs/lib/libcthyb_c.so)
==11059== by 0x1A2866E6: cthyb::sorted_spaces::sorted_spaces(triqs::utility::many_body_operator const&, triqs::hilbert_space::fundamental_operator_set const&) (in /home/hoeppner/src/TRIQS/install_triqs/lib/libcthyb_c.so)
==11059== by 0x1A202314: cthyb::solver_core::solve(cthyb::solve_parameters_t const&) (in /home/hoeppner/src/TRIQS/install_triqs/lib/libcthyb_c.so)
==11059== by 0x19F3E727: SolverCore_solve(object, object, _object*) (in /home/hoeppner/src/triqs_1.2/INSTALL/lib/python2.7/site-packages/pytriqs/applications/impurity_solvers/cthyb/cthyb.so)
==11059== by 0x4EDEF15: PyObject_Call (in /usr/lib64
...
<<<<<<<<<<<<<<< BT <<<<<<<<<<<<<<<<<<

However, if the shift in the k-mesh is not present, or we don't use the auto-partitioning scheme (partition_method = 'none'), the calculation runs smoothly. With other one-particle Hamiltonians and a small shift in the k-mesh, the calculations are running, too.

I attached a plot of the diagonal elements of G0

and plot of the differences of all elements of G0 with / without an offset of the k-mesh:

As you can the see, the differences in G0 are tiny. I appreciate any helpful comment.

All the best,
Marc

I compiled the current version of cthyb, and now I get the following error while trying to run a script:

Traceback (most recent call last):
File "Fe-hcp.py", line 119, in
S.solve(h_loc=Hloc, *_p)
File "/home/x_leopo/TRIQS1.2/install/lib/python2.7/site-packages/pytriqs/applications/impurity_solvers/cthyb/cthyb_solver.py", line 90, in solve
SolverCore.solve(self, h_loc = h_loc, *_params_kw)
TypeError:
---- There are 2 errors in Python -> C++ transcription for the class solve_parameters_t

1 The parameter 'fit_min_n' is not recognized.
2 The parameter 'fit_max_n' is not recognized.

This script was passing fine about a month ago. There I put these tail fitting parameters into p dict:
p["fit_min_n"] = fit_start
p["fit_max_n"] = fit_start+fit_interval

and **p is then passed to S.Solve

Seen on Ubuntu 14.04 machines:

/usr/lib/openmpi/include/mpi_portable_platform.h:374:63: warning: invalid suffix on literal; C++11 requires a space between literal and string macro [-Wliteral-suffix]
              _STRINGIFY(__GNUC__)"."_STRINGIFY(__GNUC_MINOR__)"."_STRINGIFY(__GNUC_PATCHLEVEL__)

I calculated single site Bethe by old and new solver with same parameters. Results for self-energy are shown in picuture.

Parameters:
U = 3.0
half_bandwidth = 0.4
chemical_potential = U/2.0+0.2
beta = 40.0
n_cycles = 100000
length_cycle = 200
n_warmup_cycles = 10000
n_l=30

Also I tried n_l=50, results are similar.

I used old version of solver (download at 25 august 2014) and solver 1.2-dev (download at 9 october 2014). Following script for old version:

from pytriqs.gf.local import *
from pytriqs.operators import *
from pytriqs.archive import *
import pytriqs.utility.mpi as mpi

# Set up a few parameters
U = 3.0
half_bandwidth = 0.4
chemical_potential = U/2.0+0.2
beta = 40.0
n_loops = 1

n_cycles = 100000
length_cycle = 200
n_warmup_cycles = 10000

# Construct the CTQMC solver
from pytriqs.applications.impurity_solvers.cthyb_matrix import Solver
S = Solver(beta = beta, gf_struct = [ ('up',[1]), ('down',[1]) ])

# Initalize the Green's function to a semi circular
S.G <<= SemiCircular(half_bandwidth)

 # Now do the DMFT loop
 for IterationNumber in range(n_loops):

    # Compute S.G0 with the self-consistency condition while imposing paramagnetism
    g = 0.5 * ( S.G['up'] + S.G['down'] )
    for name, g0block in S.G0:
        g0block <<= inverse( iOmega_n + chemical_potential - (half_bandwidth/2.0)**2  * g )

    # Run the solver
    S.solve(H_local = U * N('up',1) * N('down',1),                          # Local Hamiltonian
            quantum_numbers = { 'Nup' : N('up',1), 'Ndown' : N('down',1) }, # Quantum Numbers (operators commuting with H_Local)
        n_cycles = n_cycles,                                            # Number of QMC cycles
        length_cycle = length_cycle,                                    # Length of a cycle
        n_warmup_cycles = n_warmup_cycles,                              # How many warmup cycles
        n_legendre = 30,                                                # Use 30 Legendre coefficients to represent G(tau)
        random_name = "mt19937",                                        # Use the Mersenne Twister 19937 random generator
        use_segment_picture = False)                                    # Here we can use the segment picture

    # Some intermediate saves
    if mpi.is_master_node():
      R = HDFArchive("single_site_bethe.h5")
      R["G-%s"%IterationNumber] = S.G
      del R

And for new version:

from pytriqs.archive import HDFArchive
import pytriqs.utility.mpi as mpi
from pytriqs.gf.local import *
from pytriqs.operators.operators2 import *
from pytriqs.applications.impurity_solvers.cthyb import *

#  Example of DMFT single site solution with CTQMC
# set up a few parameters
half_bandwidth = 0.4
U = 3.0
mu = (U/2.0)+0.2
beta = 40.0
n_loops=1

# Parameters
p = {}
p["verbosity"] = 3
p["n_cycles"] = 100000
p["length_cycle"] = 200
p["n_warmup_cycles"] = 10000
p["measure_g_l"] = True
p["measure_g_tau"] = False

gf_struct = {'up':[0],'down':[0]}

# Construct solver
S = Solver(beta=beta, gf_struct=gf_struct, n_l=30)

# Local Hamiltonian
H = U*n("up",0)*n("down",0)

# init the Green function
S.G_iw <<= SemiCircular(half_bandwidth)

# Now do the DMFT loop
for IterationNumber in range(n_loops):
    # Impose Paramagnetism
    gpara = 0.5*(S.G_iw['up']+S.G_iw['down'])
    for name, g in S.G_iw: g <<= gpara

    # Compute G0
    for name, g0 in S.G0_iw:
      g0 <<= inverse( iOmega_n + mu - (half_bandwidth/2.0)**2  * S.G_iw[name] )
    S.solve(h_loc=H, **p)

# Calculation is done. Now save a few things
Results = HDFArchive("single_site_bethe.output.h5",'w')
Results["Sigma_iw"] = S.Sigma_iw
Results["G_tau"] = S.G_tau
Results["G_iw"] = S.G_iw
Results["G_l"] = S.G_l

Hello,

I get this error message

Warning : The TRIQS_HDF5_data_scheme AtomicDiagonalization is not recognized. Returning as a group. Hint : did you forgot to import this python class ?

when I try to load an AtomicDiag object from an archive. The Python class is imported as requested,

from pytriqs.applications.impurity_solvers.cthyb import AtomDiag

Is this because the data scheme name does not match the Python class name?

Hi!
I have faced following error in most recent CTHYB(downloaded 5 days back). Same input file is working fine with the previous one.

S.solve(h_loc=h_loc, **p)
File "/home/soumen/INSTALLPACKAGES/15_05_15/lib/python2.7/dist-packages/pytriqs

/applications/impurity_solvers/cthyb/cthyb_solver.py", line 90, in solve
SolverCore.solve(self, h_loc = h_loc, **params_kw)
TypeError:
---- There are 3 errors in Python -> C++ transcription for the class solve_parameters_t

1 The parameter 'fit_min_n' is not recognized.
2 The parameter 'fit_n_moments' is not recognized.
3 The parameter 'fit_max_n' is not recognized.

Thnaks,

cthyb's interface has recently undergone a major overhaul. In particular, Delta(\tau) and G(\tau) are now fixed to use the same number of imaginary time slices.
I my opinion, this can be inconvenient in many cases. Normally I take a large number of tau-points in Delta (~10000) to ensure absence of interpolation errors. But G(\tau) has to be sampled on a much sparser mesh to make binning efficient.
While having only one parameter n_tau is not necessarily bad (less input parameters = clearer interface), it would be nice to have a tool to downsample G(\tau) to a sparser mesh.

An object gf_struct, passed to the constructor of solver_core must support string-valued internal indices. This is needed for conformance with functions in pytriqs.operators.hamiltonians

There is no 1.2 tag in the cthyb tree. Since there were things moved around quite a lot, the cthyb HEAD does not compile with the tagged TRIQS 1.2 version. Can we add a 1.2 tag for CTHYB that is consistent with TRIQS 1.2? Otherwise it is difficult to maintain a stable productive version for calculations.

With the most recent version of cthyb almost all tests fail on weiss:

Total Test time (real) = 2.38 sec

The following tests FAILED:
5 - anderson (Failed)
6 - anderson_block (Failed)
7 - anderson_qn (Failed)
8 - anderson_block_qn (Failed)
9 - spinless (Failed)
10 - spinless_qn (Failed)
11 - kanamori (Failed)
12 - kanamori_qn (Failed)
13 - kanamori_offdiag (Failed)
14 - kanamori_offdiag_qn (Failed)
15 - legendre (Failed)
16 - atomic_gf (Failed)
17 - kanamori (Failed)
18 - single_site_bethe (Failed)
19 - slater (Failed)

the error is

Starting on 1 Nodes at : 2014-11-18 19:28:41.280547
Traceback (most recent call last):
File "", line 5, in
File "/home/leonid/TRIQS1.2/build_cthyb/pytriqs/applications/impurity_solvers/cthyb/init.py", line 26, in
from cthyb_solver import Solver
File "/home/leonid/TRIQS1.2/build_cthyb/pytriqs/applications/impurity_solvers/cthyb/cthyb_solver.py", line 1, in
from cthyb import SolverCore
ImportError: /home/leonid/TRIQS1.2/build_cthyb/pytriqs/applications/impurity_solvers/cthyb/cthyb.so: undefined symbol: _ZNK5cthyb13sorted_spaces9atomic_gfEdRKSt3mapISsSt6vectorIN5boost7variantIiSsNS3_6detail7variant5void_ES7_S7_S7_S7_S7_S7_S7_S7_S7_S7_S7_S7_S7_S7_S7_S7_S7_EESaIS8_EESt4lessISsESaISt4pairIKSsSA_EEEi

• One page about all officially supported moves with short descriptions.
• One page about measurements with definitions of the measured quantities.

After running cmake (in a non-empty build directory), I get something like this:

-- Found TRIQS: /home/pseth/SOURCES/TRIQS/install/lib/libtriqs.so
CMake Error at python/CMakeLists.txt:8 (triqs_cython_module):
Unknown CMake command "triqs_cython_module".

-- Configuring incomplete, errors occurred!
make: *** No rule to make target `install'. Stop.

This is resolved with "cmake .", hinting to a possible error in cmake files.

hi just managed to build TRIQS 1.2 on our cluster,
as one of our users requests cthyb_matrix I checked it out and ran cmake.
Unknown CMake command "triqs_cython_module".
cmake -DTRIQS_PATH=$EBROOTTRIQS ../cthyb_matrix
-- TRIQS : Adding compilation flags detected by the library (C++11/14, libc++, etc...)
CMake Error at python/CMakeLists.txt:8 (triqs_cython_module):
Unknown CMake command "triqs_cython_module".

-- Configuring incomplete, errors occurred!
See also "*********************/scratch/triqs-1.2/build/CMakeFiles/CMakeOutput.log".

this is mentioned in issue #4
any help welcome
does git master branch correspond to version 1.2?

Currently the plots in the doc are not working due to some sphinx issue with the .. plot command. I believe Olivier looked at this and knows where the problem lies. I am putting it up here before I forget of the existence of this problem!

Hello!

If I want to run a DMFT calculation using the CTHYB segment picture solver, what do I need to do to when I instantiate the solver class?

Regards
Johan Jönsson

Dear TRIQS team,

I try to install the cthyb on my desktop, but it shows the following error:

Scanning dependencies of target cthyb_c
[ 4%] Building CXX object c++/CMakeFiles/cthyb_c.dir/solver_core.cpp.o
[ 8%] Building CXX object c++/CMakeFiles/cthyb_c.dir/atom_diag.cpp.o
[ 13%] Building CXX object c++/CMakeFiles/cthyb_c.dir/atom_diag_functions.cpp.o
[ 17%] Building CXX object c++/CMakeFiles/cthyb_c.dir/atom_diag_worker.cpp.o
[ 21%] Building CXX object c++/CMakeFiles/cthyb_c.dir/impurity_trace.cpp.o
[ 26%] Building CXX object c++/CMakeFiles/cthyb_c.dir/measure_density_matrix.cpp.o
Linking CXX shared library libcthyb_c.so
[ 26%] Built target cthyb_c
[ 30%] Generating cthyb/cthyb_wrap.cpp, ../include/cthyb/py_converters/cthyb.hpp
/usr/bin/python: can't open file '/home/thlee/library/cthyb/build/python/cthyb/cthyb_desc.py': [Errno 2] No such file or directory
make[2]: *** [python/cthyb/cthyb_wrap.cpp] Error 2
make[1]: *** [python/CMakeFiles/cthyb.dir/all] Error 2
make: *** [all] Error 2

I'm able to install cthyb on my laptop, and I compare the difference between my laptop and desktop. I found that there's no such directory "/build/python/cthyb/" on my laptop(which I successfully installed cthyb). Also, there's a file "sitecustomize.py" on my laptop in the build top directory, but there's no such file on my desktop. I'm wondering how should I solve this problem?

Thank you.
Tsung-Han

I was caught by some unexpected though explained behaviour of the solve method. In trying to accumulate some statistics on a converged calculation, I noticed that S.G0_iw was slightly different at each measurement.

In the script below, G0_iw before the solve and after are different as the 1/iw behaviour is changed in the solve routine, see solver_core.cpp around line 121.

S.G0_iw << S.Sigma_iw + inverse(S.G_iw)
S.G0_iw << inverse(S.G0_iw)
ar = HDFArchive('testfile.h5','a')
ar['G0_iw-orig'] = S.G0_iw
del ar

# Start DMFT iterations:
for iteration_number in range(1,loops+1):
    ar = HDFArchive('testfile.h5','a')
    ar['G0_iw-before-%s'%(iteration_number)] = S.G0_iw
    del ar

    p["random_seed"] = 34788 + 928374*mpi.rank + iteration_number + previous_runs
    # Solve the impurity problem:
    S.solve(h_loc=h_loc, **p)

    # Write the final Sigma and G to the hdf5 archive:
    ar = HDFArchive('testfile.h5','a')
    ar['G0_iw-after-%s'%(iteration_number)] = S.G0_iw
    ar['G_tau-%s'%(iteration_number)] = S.G_tau
    del ar

Do we want to correct the 1/iw behaviour in G0_iw? I did this as it was what was done in the previous solver, but I am not sure it is necessary or good. The user probably expects the input to not change.

We need to check if git is actually being used for the repo. Otherwise, these lines in src/CMakeLists.txt throw an error.

# Get hash
triqs_get_git_hash(${CMAKE_SOURCE_DIR} "CTHYB")
message(STATUS "Hash: ${CTHYB_GIT_HASH}")

After compilation, succesful tests, and installation, one cannot load the installed module. The reason is that in the installation path, there is no init.py in pytriqs/applications! Adding an empty init.py file at pytriqs/applications and pytriqs/applications/impurity_solvers resolves this problem.
BTW, for the same reason all the moduls in applications/dft are also not loadable!

Just a reminder.

There are rare events, when insertion of a new operator fails in move_shift_operator::attempt(). Such events result in an immediate solver crash:

 Traceback (most recent call last):
  File "./5_plus_5.py", line 100, in <module>
    S.solve(**p)
RuntimeError: Error occurred at Mon Nov  3 22:53:17 2014

Error calling C++ overload
 solve() -> void
in implementation of method SolverCore.solve
C++ error was :
 Triqs runtime error at
/home/igor/Physics/TRIQS/cthyb.git/c++/triqs/utility/rbt.hpp : 183

 node 12.2577 [time_pt : beta = 40 n = 5652854765438934592] is flagged for delete
 Error occurred on node 0
 C++ trace is :
/home/igor/Physics/TRIQS/installed/lib64/python2.7/site-packages/pytriqs/applications/impurity_solvers/cthyb/cthyb.so
triqs::exception::exception() 0x43 [0x7fac4dfd7fb3]
 /home/igor/Physics/TRIQS/installed/lib/libcthyb_c.so
rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t, std::greater<triqs::utility::time_pt>
>::clear_modified_impl(rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t,
std::greater<triqs::utility::time_pt> >::node_t*) 0x88
[0x7fac4dd1eef8]
 /home/igor/Physics/TRIQS/installed/lib/libcthyb_c.so
rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t, std::greater<triqs::utility::time_pt>
>::clear_modified_impl(rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t,
std::greater<triqs::utility::time_pt> >::node_t*) 0x4b
[0x7fac4dd1eebb]
 /home/igor/Physics/TRIQS/installed/lib/libcthyb_c.so
rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t, std::greater<triqs::utility::time_pt>
>::clear_modified_impl(rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t,
std::greater<triqs::utility::time_pt> >::node_t*) 0x5d
[0x7fac4dd1eecd]
 /home/igor/Physics/TRIQS/installed/lib/libcthyb_c.so
rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t, std::greater<triqs::utility::time_pt>
>::clear_modified_impl(rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t, std::greater<triqs::utility::time_pt> >::node_t*) 0x5d
[0x7fac4dd1eecd]
 /home/igor/Physics/TRIQS/installed/lib/libcthyb_c.so
rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t, std::greater<triqs::utility::time_pt>
>::clear_modified_impl(rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t,
std::greater<triqs::utility::time_pt> >::node_t*) 0x4b
[0x7fac4dd1eebb]
 /home/igor/Physics/TRIQS/installed/lib/libcthyb_c.so
rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t, std::greater<triqs::utility::time_pt>
>::clear_modified_impl(rb_tree<triqs::utility::time_pt, cthyb::impurity_trace::node_data_t, std::greater<triqs::utility::time_pt> >::node_t*) 0x5d
[0x7fac4dd1eecd]
 /home/igor/Physics/TRIQS/installed/lib/libcthyb_c.so
cthyb::move_shift_operator::attempt() 0xc72 [0x7fac4dd374b2]
 /home/igor/Physics/TRIQS/installed/lib/libcthyb_c.so
triqs::mc_tools::move_set<double>::attempt() 0xa2 [0x7fac4dd0b032]
 /home/igor/Physics/TRIQS/installed/lib/libcthyb_c.so
triqs::mc_tools::Step::Metropolis<double>::do_it(triqs::mc_tools::move_set<double>&, triqs::mc_tools::random_generator&, double&) 0x17 [0x7fac4dd0aed7]
 /home/igor/Physics/TRIQS/installed/lib/libcthyb_c.so
triqs::mc_tools::mc_generic<double,
triqs::mc_tools::Step::Metropolis<double> >::start(double, std::function<bool ()> const&) 0x15e [0x7fac4dd0466e]
 /home/igor/Physics/TRIQS/installed/lib/libcthyb_c.so
cthyb::solver_core::solve(cthyb::solve_parameters_t const&) 0x2d0a
[0x7fac4dcf9d9a]
 /home/igor/Physics/TRIQS/installed/lib64/python2.7/site-packages/pytriqs/applications/impurity_solvers/cthyb/cthyb.so
0xe996 [0x7fac4dfd5996]
 /usr/lib64/libpython2.7.so.1.0 PyEval_EvalFrameEx 0x4333 [0x7fac616fa1b3]
 /usr/lib64/libpython2.7.so.1.0 PyEval_EvalCodeEx 0x7dd [0x7fac616fb99d]
 /usr/lib64/libpython2.7.so.1.0 PyEval_EvalCode 0x32 [0x7fac616fbaa2]
 /usr/lib64/libpython2.7.so.1.0 0xfa44f [0x7fac6171444f]
 /usr/lib64/libpython2.7.so.1.0 PyRun_FileExFlags 0x7e [0x7fac6171556e]
 /usr/lib64/libpython2.7.so.1.0 PyRun_SimpleFileExFlags 0xd9 [0x7fac617166d9]
 /usr/lib64/libpython2.7.so.1.0 Py_Main 0xcb3 [0x7fac61727183]
 /lib64/libc.so.6 __libc_start_main 0xf5 [0x7fac61079dc5]
 /usr/bin/python2.7 [0x4009fd]

I've found that the crash goes away, if I comment out data.imp_trace.cancel_insert(); in this block of code:

  // Try to insert the new operator at shifted time in the tree
  try {
   data.imp_trace.try_insert(tau_new, op_new);
  }
  catch (rbt_insert_error const&) {
   //std::cerr << "Insert error : recovering ... " << std::endl;
   data.imp_trace.cancel_insert();
   return 0;
  }

I get the following error message while trying to install new cthyb on weiss:

CMake Error at python/CMakeLists.txt:12 (triqs_python_extension):
Unknown CMake command "triqs_python_extension".

triqs1.2 core was compiled and installed without any problems

Hi!
Folling is not working with new version :::

R=HDFArchive(dft_filename+'.h5','a')
File "/home/soumen/INSTALLPACKAGES/15_05_15/lib/python2.7/dist-packages/pytriqs/archive/hdf_archive.py", line 377, in init
self._init_root( LocalFileName, open_flag)
File "/home/soumen/INSTALLPACKAGES/15_05_15/lib/python2.7/dist-packages/pytriqs/archive/hdf_archive_basic_layer_h5py.py", line 38, in _init_root
fich = h5py.File(LocalFileName, open_flag)
File "/usr/lib/python2.7/dist-packages/h5py/_hl/files.py", line 207, in init
fid = make_fid(name, mode, userblock_size, fapl)
File "/usr/lib/python2.7/dist-packages/h5py/_hl/files.py", line 98, in make_fid
fid = h5f.create(name, h5f.ACC_EXCL, fapl=fapl, fcpl=fcpl)
File "h5f.pyx", line 90, in h5py.h5f.create (h5py/h5f.c:1984)
IOError: unable to create file (File accessibilty: Unable to open file)

Thanks,

Hi,
On Ubuntu 12.04, clang 3.4, with latex version of TRIQS, I get the following error message during compilation:

https://gist.github.com/tayral/7eb624a9f7f82e879b38

Do you have any idea how to fix this?

Thanks!
Thomas

I recently compiled Triqs gcc6 branch, and cthyb master.
error is in single_site_bethe test:

Starting on 1 Nodes at : 2016-05-23 09:50:51.299382
Traceback (most recent call last):
  File "<stdin>", line 57, in <module>
RuntimeError: .. Error occurred at Mon May 23 09:51:21 2016

.. Error .. calling C++ overload 
.. assert_block_gfs_are_close(block_gf_view<imfreq> x, block_gf_view<imfreq> y, double precision) -> void 
.. in implementation of function comparison_tests.assert_block_gfs_are_close
.. C++ error was : 
.. Triqs runtime error at /home/oscar/dev/triqs/triqs/gfs/functions/gf_tests.hpp : 49

Gfs have different data. max(abs(x-y)) = 301.922
.. Error occurred on node 0