3
0
mirror of https://github.com/triqs/dft_tools synced 2024-10-31 19:23:45 +01:00
Commit Graph

517 Commits

Author SHA1 Message Date
Olivier Parcollet
41caa7ab91 Fix #139
- Python legacy code :
- Put back for both refreq and imfreq gf
g += matrix
g += complex
g + matrix
g + complex
and same with - operator.

- same impl. for both.
- with python wrapping
2014-10-22 21:19:05 +02:00
Olivier Parcollet
eacf3a0d30 Fix #77. 2014-10-22 21:19:05 +02:00
Olivier Parcollet
9688154b0b [mpi] Clean, rm mpi.hpp, Fix #133, Fix #140
- no need to include mpi, it is automatic from the array, gf containers.
- Fix example from issue 133.
- remove old example code.
2014-10-22 15:33:13 +02:00
Olivier Parcollet
ee42e24fba Fix previous commit for c++11 2014-10-21 23:49:57 +02:00
Olivier Parcollet
a34f23580a [mpi] Work on draft
- add invoke2, works now for block_gf
- TODO: more systematic tests
- TODO: remove the mpi.hpp, include as higher heder like in gf
2014-10-21 23:30:44 +02:00
Olivier Parcollet
6772481f71 [doc] Install on Ubuntu 14.04 2014-10-21 23:19:45 +02:00
Hartmut Hafermann
2705353f1d API change in mpi: rename allreduce -> all_reduce 2014-10-21 16:15:44 +02:00
Hartmut Hafermann
f4220efaea fix C++11 issue in clef 2014-10-21 13:31:32 +02:00
Olivier Parcollet
5a0efd5215 Correction 2014-10-20 17:27:06 +02:00
Olivier Parcollet
28ec8a0fdd [gfs] Various fixes, for paper 2014-10-20 17:09:56 +02:00
Olivier Parcollet
32500df5ba [gfs] Change bz tag to using the actual brillouin_zone class
- simpler to remember ...
- in this case, gf is specialized on the actual name of the domain.
2014-10-20 16:29:50 +02:00
Olivier Parcollet
21a0587ba9 [cmake] Add USE_CPP14 is TRIQS_Config
- for usage in applications
2014-10-20 16:27:36 +02:00
Olivier Parcollet
68d909f356 [doc] update compiler requirements. 2014-10-19 12:03:20 +02:00
Olivier Parcollet
46fd9a5d0e Fix for C++11 and gcc 4.8 2014-10-19 10:51:10 +02:00
Olivier Parcollet
edc059e56a [mctools] Some cleaning.
-TODO: remove parameters
- remove access to type erased move and measure...
2014-10-18 21:26:20 +02:00
Olivier Parcollet
a857aca301 Silence warnings on gcc
gcc 4.9 with -Wall -Wno-unknown-pragmas -Wno-maybe-uninitialized -Wno-sign-compare

-> uninit : not reliable apparently.
-> sign compare : to be fixed later.
-> fix all, but parenthesis suggestion
2014-10-18 21:20:42 +02:00
Olivier Parcollet
0a3f07a6cb [doc] example correction after tail changes. 2014-10-18 21:20:42 +02:00
Olivier Parcollet
a383191789 Fix one test in gcc 2014-10-18 21:20:41 +02:00
Olivier Parcollet
5eaab18b4f [cmake] Minor modif. C++14 in definition for apps.
- Change the message for cmake of apps (was misleading C++11 while 14).
- minor cleaning
2014-10-18 21:20:41 +02:00
Olivier Parcollet
3893c41fc0 Fix C++11 compilation.
- one return type deduction
2014-10-18 21:20:41 +02:00
Olivier Parcollet
ed6379ce63 [clang tool] add parameter class treatment and h5 generator
wrapper generator: add treatment of parameter class

- methods taking a parameter class are called by **kw
- the dict is passed to the C++
- the converters for the parameters are generated.
2014-10-18 21:20:41 +02:00
Olivier Parcollet
0a1285405c [gfs] Lattice fourier, multivar G, curry, tail
- Add Fourier for lattice.
  - Add regular_bz_mesh, cyclic_lattice, and their FFT.

- rm freq_infty.
- The gf can now be evaluated on a tail_view, which result in composing the tail.
- Fix the following issue :
  g(om_) << g(om_ +1)
  will recompose the tail correctly.
- TODO : TEST THIS NEW FEATURE IN DETAIL.

- Work on singularity for G(x, omega)

 - Separate the factory for singularity from the data factory in gf.
 - overload assign_from_functoin (renamed).
 - Fix singularity_t and co in the gf (const issue).

- Clean tail, add tail_const_view
 - add m_tail for x -> tail on any mesh
 - test curry + fourier works on k
2014-10-18 21:20:35 +02:00
Olivier Parcollet
47cb8a03f7 [arrays] Important changes in implementation.
- Simplify group_indices
  - Only for C ordered, remove complex compile time.
  - Could be generalized to non C ordered, but no need.
- Fix slice for custom orders.
- Generalize the group_indices for the custom order.
- Add c_ordered_transposed_view (useful ?)
- Improve slice, special for ellipsis (quicker).
- Simplify TraversalOrder
- Assignement. Specialize one case for speed.
- use FORCEINLINE in foreach, according to speed test for clang
- add one speed test
- Modify iterators for better speed.
- along the lines decided for the foreach
- update doc.
2014-10-18 21:20:17 +02:00
Olivier Parcollet
d1c8a9a46e gcc compilation detail 2014-10-18 21:20:17 +02:00
Olivier Parcollet
672834fd68 A few parts of the lib only in C++14 mode
- multivariable gf
- some part of tuple tools.
- modified tests
2014-10-18 21:20:17 +02:00
Olivier Parcollet
f8aef6a290 [gfs] Add memory layout option 2014-10-18 21:20:17 +02:00
Olivier Parcollet
fa07abbea9 [gfs] Singularity as an explicit template parameter
- change the underlying data : do not flatten the linear indices of the
  mesh into a single index, keep a higher dim array.
- easier for various places, and necessary for g(nu,nu').
- work on several 2 part. containers.
- add default target (may not be always matrix_valued)
2014-10-18 21:20:16 +02:00
Olivier Parcollet
9c129cb224 [mpi] draft of gf support
- done Matsubara freq for testing and rereading.
- TODO: generalize to other meshes.
- draft for multi var gf
2014-10-18 21:20:01 +02:00
Hartmut Hafermann
ebbb2f0b25 [mpi] vector, test and draft of doc 2014-10-18 21:19:52 +02:00
Olivier Parcollet
38cfef4e9f Implement mpi lib (1). Array, generic, base, vector
- Implement the basic structure of the mpi lib
  and specialization for arrays, basic types, std::vector
- adapted the array class for the lazy mpi mechanism
- pass tests on arrays :
   - scatter, gather on array<long,2> array<complex,2>, etc...
   - broadcast
- several files for readibility
- the std::vector coded but not tested.
- generic mecanism implemented and tested (mpi_generic test)
- added several tests for the mpi lib.
- TODO : more tests, doc...
2014-10-18 21:19:51 +02:00
Olivier Parcollet
56820a9493 [doc] install and exceptions 2014-10-18 21:19:51 +02:00
Olivier Parcollet
dfe1948d38 [cmake] Test log compare command ready to copy-paste.
- remove the ; !
2014-10-18 21:19:51 +02:00
Olivier Parcollet
dddae19853 Clean old boost and old mc_tools code (v1) 2014-10-18 21:05:07 +02:00
tayral
edd1ff4529 Restructuring documentation.
A first general restructuration of the doc according to the pattern [tour|tutorial|reference].
In the reference part, objects are documented per topic.
In each topic, [definition|c++|python|hdf5] (not yet implemented)
2014-10-18 12:21:08 +01:00
tayral
a9e5f20c39 put back literalinclude instead of triqs_example for c++ 2014-10-17 17:01:20 +01:00
tayral
880f30b086 Changed all literalincludes --> runblock / triqs_example
This is to avoid keeping code snippets that do not work in the doc. At least there will be an error message.
2014-10-17 16:50:46 +01:00
Priyanka Seth
70d4aba545 Determining delta from gf (currently only BlockGf and GfImFreq) 2014-10-17 15:55:03 +02:00
Priyanka Seth
ec32223ea7 Clean up _gf_common.py -- removed duplicated python functions 2014-10-17 10:21:32 +02:00
Olivier Parcollet
c1f221e085 [arrays] More include by default 2014-10-16 22:32:11 +02:00
Olivier Parcollet
5b6b0ecf38 Revert "include map and fold headers into array.hpp"
This is not the right place for include...

This reverts commit 63fd1b8b2d.
2014-10-16 22:21:37 +02:00
Hartmut Hafermann
63fd1b8b2d include map and fold headers into array.hpp 2014-10-16 16:30:48 +02:00
Priyanka Seth
d8e79056bf Modify det_manip to access determinant x_values and y_values 2014-10-14 16:26:08 +02:00
Hartmut Hafermann
e93adc0b84 [gfs] add << operator for gfs
kept <<= for backward compatibility
TODO: clean doc and examples, possibly add deprecation warning
2014-10-13 17:42:11 +02:00
Michel Ferrero
2f2374ea29 Fix issue #128
This is just cosmetics but should be fine.
2014-10-09 13:15:58 +02:00
Michel Ferrero
173a97d07c Fix for issue #134
There was a bug in the transpose of the indices. Added a test.
2014-10-09 12:46:07 +02:00
Olivier Parcollet
0fd82579e3 [cmake] Fix cmake python for Ubuntu 14.04
There is a bug in Ubuntu 14.04 : the interpreter
does not return the correct position of the libpython.
We add the correct path to the path hints, at the end (lowest priority).
In most machines, the first hint is the correct answer.
2014-10-08 17:01:40 +02:00
Priyanka Seth
6a38d6747c Fix in tail-matrix multiplication
Set order_max to correspond to the mask
2014-10-07 13:55:56 +02:00
Priyanka Seth
9be131e8e1 Allow indices in gf_struct/operator names to be strings. 2014-10-07 13:53:18 +02:00
Michel Ferrero
3b59d26a67 Two fixes in tight binding part 2014-10-06 23:18:10 +02:00
Priyanka Seth
71fa498833 API change for tail fitters
python fit_tail, replace_by_tail ==> fit_tail_depr, replace_by_tail_depr
c++ set_tail_from_fit ==> fit_tail
2014-10-01 10:52:32 +02:00