Code for reduction of Near-Field and Far-Field High Energy Diffraction Microscopy (HEDM) data.

Neldermead is taken from http://people.sc.fsu.edu/~jburkardt/cpp_src/asa047/asa047.html and modified to include constraints and use as CUDA kernels. SGInfo library used to calculate HKLs. Need to install libtiff-dev and nlopt for compilation of NF-HEDM codes.

To check help for installation, type "make help" in the terminal. For individual help type "make helpnf" or "make helpff" in the terminal. To compile individually, need to go to the sub-folder and "make" individually. Would need NLOPT and TIFF packages. For experimental CUDA codes: go to FF_HEDM folder and "make cuda". This doesn't require any external library.

Stampede is a part of the XSEDE supercomputing network. MIDAS far-field HEDM (FF) installation is tested on Stampede. To install MIDAS FF code on Stampede, Log into Stampede and change to the Home directory.
cd
Clone MIDAS Github repo.
git clone https://github.com/hmparanjape/MIDAS
Change to the MIDAS FF folder
cd MIDAS/FF_HEDM
Load NetCDF library which is required by MIDAS
module load netcdf
Run the make command to compile binaries and create necessary scripts. This may take several minutes.
make stampede
In the end, you should see the following success message
Congratulations, you can now use MIDAS on Stampede to run FarField analysis

In the Example.Stampede there is a sample parameter file and a job file to run MIDAS. MIDAS jobs need to run on a largmem (large memory) node. Create a folder in the WORK directory to run a simulation. Copy MIDAS FF sample files, modify them and submit a job.
cdw
mkdir midas_test; cs midas_test
cp ~/MIDAS/FF_HEDM/Example.Stampede/* .
sbatch midas_test.job

To install on a local computer, go to FF_HEDM folder and "make local". This will download NLOPT and SWIFT packages and install shortcuts in ${HOME}/.MIDAS directory.