MCD-PYTHON: python-based interface to the Mars Climate Database

Open source code and contact information available on github [no registration needed]

• To get sources through git

  git clone https://github.com/aymeric-spiga/mcd-python

• To get sources through SVN

  svn co https://github.com/aymeric-spiga/mcd-python/trunk mcd-python

• To get a static ZIP file of the current version of the code, click here

How to install?

It is assumed you were able to compile successfully the mcd sources with gfortran.

The netCDF library should be installed on your system. Moreover, from our experience, it seems that it must have been build using the -fPIC (for gfortran; the name of the option changes with compilers) option which generates position independent code suitable for use in a shared library. An example script is given in the netcdf folder.

The installation below relies on f2py utility, which is part of the numpy package.

1. Getting the environment variables right: add the mcd-python folder to PYTHONPATH in your environment file (e.g. .bashrc)

   export PYTHONPATH=$PYTHONPATH:adapt_to_your_own/mcd-python

2. Modify the compile script compile_fmcd.sh to link your local netCDF libraries and mcd distribution (Fortran sources)

3. Check that f2py is included in your python library suite.

4. Run compile_fmcd.sh and check for the created .so file (its size should be about 1 Mo)

Quick test

quicktest.py

Next step is to try and learn about the use of mcd Python library with the tutorial folder.

A more advanced example (direct use of fmcd compiled with f2py) is provided in the test_mcd folder.

Python 3

Solution suggested by Aaron Berliner

This can be done using the 2to3 package and the reindent

Run 2to3 -v -n -W -f all mcd.py
Run 2to3 -v -n -W -f all mcdcomp.py
Run reindent mcd.py
Run reindent mcdcomp.py

Then upgrade to the appropriate basemap in python3.