-
Notifications
You must be signed in to change notification settings - Fork 14
Installation
ALPSCore needs to be properly installed, see ALPSCore library.
Only header-file libraries are needed. The dependencies will be taken care of by ALPSCore.
Head-file libraries for linear algebra.
MPI environment is required.
- Mac OS X
To be written
The solver depends on ALPSCore libraries and some Boost libraries (timer, chrono, system). These libraries must be preinstalled. Note that the Boost timer, chrono, system libraries are not header-file-only libraries and must be compiled. If you want to install the TRIQS-compatible Python interface, please follow the procedure described in another website.
The CT-HYB package can be obtained by following methods:
- Clone Git repository at Github
$ git clone https://github.com/ALPSCore/CT-HYB.git
- From release tarball:
-
Download a release tarball from https://github.com/ALPSCore/CT-HYB/releases
-
Rename the downloaded tarball to CT-HYB.tar.gz (or CT-HYB.zip, if you chose the zip version) and unpack it:
$ tar -xzf CT-HYB.tar.gz
-
Then, make a (separated) build directly, and provide something like:
$ mkdir build
$ cd build
$ cmake\
$ -DALPSCore_DIR=/path/to/ALPSCore \
$ -DCMAKE_INSTALL_PREFIX=/path/to/install/dir \
$ -DCMAKE_CXX_COMPILER=/path/to/C++/compiler \
$ ../CT-HYB
$ make
$ make test
$ make install
You must use a MPI C++ compiler. This may be done by setting the path of your MPI wrapper compiler (such as mpic++) to CMAKE_CXX_COMPILER. Note that, in such cases, MPI must be enabled also in the installation of ALPSCore. If cmake does not find boost, please tell cmake the installation directory of boost by using the option "-DBOOST_ROOT=***".
We describe advanced options in installation.
- Enable support for quadruple precision floating point numbers
When you solve an impurity model at very low temperature, it may be better to activate this option to avoid underflow and overflow in the evaluation of the local trace. When this option is on, the value of the trace is stored as a quadruple precision floating point number, while matrix manipulations are still performed using double precision floating point numbers. This makes the execution bit slower.
$cmake -DUSE_QUAD_PRECISION=ON ...
- Some libraries are not found at runtime.
When you install the executalbe to your installation path by "make install", CMake removes the paths of dynamic libraries from the binary. When you launch "/path/to/install/dir/hybmat", some dynamic libraries which were visible in the build may not be found. In this case, please set your environment variables correctly (e.g., LD_LIBRARY_PATH) so that the system can find these libraries at runtime. More information is found [here] (https://cmake.org/Wiki/CMake_RPATH_handling).