This thread is considered to be outdated, but is kept for reference. For the most current source code and compiling instructions, see:
https://www.mystran.com/forums/showthread.php?tid=83
LAST EDIT: 4/19/2021
-----------------
SOURCE CODE
-----------------
Official GitHub: https://github.com/dr-bill-c/MYSTRAN
Note that there are NASA links for MYSTRAN, which should now point to www.mystran.com or https://github.com/dr-bill-c/MYSTRAN. There are other GitHub forks as well, but the official and most recent version is linked above.
For official versions of the source code (source code in a Zip file as opposed to the current GitHub): https://www.mystran.com/forums/showthread.php?tid=39
Note there is a 7 page document "MYSTRAN-Source-Code-Structure.pdf" that briefly describes the source code file structure within the "Documentation" in this link: https://www.mystran.com/forums/showthread.php?tid=39
------------------------------
12.0 and 12.1 COMPILING
------------------------------
The compiling process has changed since 11.2 because the SuperLU sparse solver has been incorporated.
-----Overall Notes (Compares Options)
IN WORK
-----Option 1 (Preferred Approach): CMake + gFortran (Widows and Linux)
https://www.mystran.com/compile/MYSTRAN_CMake_12.0.pdf
The following link is the original source of the above PDF:
https://gist.github.com/Bruno02468/4b58a...f2008698f5
Note that you will use the SuperLU sparse solver, which is part of the build. The above process directly uses v5.2.2, which is linked below for reference:
https://www.mystran.com/SuperLU/superlu_5.2.2.tar.gz
For reference, the original source for SuperLU is:
https://portal.nersc.gov/project/sparse/superlu/
-----Option 2 (Alternative): Code::Blocks + gFortran (Windows Instructions)
First, you need to address the SuperLU sparse solver, which is part of the build. To download the source code, we are using the version from this link:
https://portal.nersc.gov/project/sparse/superlu/
Source code direct link:
https://www.mystran.com/SuperLU/superlu_5.2.2.tar.gz
In Windows 10, you can compile SuperLU with this procedure, which requires CMake
https://www.mystran.com/SuperLU/Compile_SuperLU.pdf
The compiled SuperLU binaries are not provided because of the chance there can be an incompatibility. It is preferred to compile the yourself with the provided instructions above.
Step 2: How to use Code::Blocks with the SuperLU files for Windows
https://www.mystran.com/compile/MYSTRAN_...s_12.0.pdf
-------------------
BENCHMARK
-------------------
In order to benchmark a new build, you can run the DAT files here and compare the resulting ANS files to the ARC files. There is a batch file that allows you perform all the runs.
We are working on an Excel file that can do a detailed comparison on the numerical data. In the meantime, you verify that the resulting ANS file sizes are the same (or very similar) to the ARC files.
https://www.mystran.com/forums/showthread.php?tid=39
-------------------------------------------------
ANSIBLE ROLE FOR INSTALLING MYSTRAN
-------------------------------------------------
https://www.mystran.com/forums/showthread.php?tid=67
https://www.mystran.com/forums/showthread.php?tid=83
LAST EDIT: 4/19/2021
-----------------
SOURCE CODE
-----------------
Official GitHub: https://github.com/dr-bill-c/MYSTRAN
Note that there are NASA links for MYSTRAN, which should now point to www.mystran.com or https://github.com/dr-bill-c/MYSTRAN. There are other GitHub forks as well, but the official and most recent version is linked above.
For official versions of the source code (source code in a Zip file as opposed to the current GitHub): https://www.mystran.com/forums/showthread.php?tid=39
Note there is a 7 page document "MYSTRAN-Source-Code-Structure.pdf" that briefly describes the source code file structure within the "Documentation" in this link: https://www.mystran.com/forums/showthread.php?tid=39
------------------------------
12.0 and 12.1 COMPILING
------------------------------
The compiling process has changed since 11.2 because the SuperLU sparse solver has been incorporated.
-----Overall Notes (Compares Options)
IN WORK
-----Option 1 (Preferred Approach): CMake + gFortran (Widows and Linux)
https://www.mystran.com/compile/MYSTRAN_CMake_12.0.pdf
The following link is the original source of the above PDF:
https://gist.github.com/Bruno02468/4b58a...f2008698f5
Note that you will use the SuperLU sparse solver, which is part of the build. The above process directly uses v5.2.2, which is linked below for reference:
https://www.mystran.com/SuperLU/superlu_5.2.2.tar.gz
For reference, the original source for SuperLU is:
https://portal.nersc.gov/project/sparse/superlu/
-----Option 2 (Alternative): Code::Blocks + gFortran (Windows Instructions)
First, you need to address the SuperLU sparse solver, which is part of the build. To download the source code, we are using the version from this link:
https://portal.nersc.gov/project/sparse/superlu/
Source code direct link:
https://www.mystran.com/SuperLU/superlu_5.2.2.tar.gz
In Windows 10, you can compile SuperLU with this procedure, which requires CMake
https://www.mystran.com/SuperLU/Compile_SuperLU.pdf
The compiled SuperLU binaries are not provided because of the chance there can be an incompatibility. It is preferred to compile the yourself with the provided instructions above.
Step 2: How to use Code::Blocks with the SuperLU files for Windows
https://www.mystran.com/compile/MYSTRAN_...s_12.0.pdf
-------------------
BENCHMARK
-------------------
In order to benchmark a new build, you can run the DAT files here and compare the resulting ANS files to the ARC files. There is a batch file that allows you perform all the runs.
We are working on an Excel file that can do a detailed comparison on the numerical data. In the meantime, you verify that the resulting ANS file sizes are the same (or very similar) to the ARC files.
https://www.mystran.com/forums/showthread.php?tid=39
-------------------------------------------------
ANSIBLE ROLE FOR INSTALLING MYSTRAN
-------------------------------------------------
https://www.mystran.com/forums/showthread.php?tid=67