SHORT INSTRUCTIONS (assumes you're running on 4 processors)

1. Link all SPECFEM3D binaries from your local installation into folder ./binary_files/, e.g.:

$ ln -s ~/SPECFEM3D/bin/* binary_files/

The workflow is compiled to work on the most recent svn version of the code without any modifications.  The programs that are needed are: 
xspecfem3D 
xdecompose_mesh_SCOTCH 
xgenerate_databases 
xcombine_vol_data

2. Change the values of the following variables (default values shown), defined in ./inversion/input_parameters.py:

data_path = '/export/data/schiemenz/workflow_examples/toy_cube/generate_data/true_model_seismograms/'
CMTSOLUTION_path = '/export/data/schiemenz/workflow_examples/toy_cube/input_files/CMTSOLUTION_files/'
event_scratch_directory='/export/data/schiemenz/workflow_examples/toy_cube/inversion/event_scratch_directory/'

... in general, you need only replace the part
/export/data/schiemenz/workflow_examples/toy_cube/

to wherever the workflow has been locally installed

3. run the script 
$ ./do_everything.sh


VISUALIZATION ROUTINES
./inversion/visualization/plot_well_log.py    -- plots a depth profile at (x,y) = (500,500)
./inversion/visualization/create_vtks.py      -- creates *.vtk files for the models for paraview visualization
./libraries/xyztovtk.py                       -- converts a tomographic model (.xyz file) into a .vtk file for paraview