Tomographic reconstruction of Lyman-alpha forest flux using a Wiener filter.
Dachshund is written in C++. You will need a C++ compiler, preferably with OpenMP support. On NERSC, make sure to use the gcc compiler rather than the default intel compiler (see http://www.nersc.gov/users/software/compilers/gnu/).
The make system is straightforward. Create the 'platform.make' file by copying the relevant file from the 'platforms' directory. The 'platform.make' file controls some system-specific settings, and there are examples in the 'platforms' directory.
The default make
target builds the dachshund library, the dachshund
application, and runs the test suite.
The dachshund application requires an input config file specifying the run-time
parameters and an input data file containing the pixel data. The config file
format might change during development, so please check against
app/dachshund.cc
(the main application source) to make sure it is up to date.
Then run the dachshund application with:
$ ./dachshund.ex input.cfg
The pixel data file should contain the spatial coordinate
Dachshund uses Eigen for some linear algebra routines and Catch for organizing tests. Both are included in the repository, so there is no need to get them separately.