Deal with cases where experiments are encoded as Python scripts

[bilalshaikh42]

@jonrkarr is this something that we would be able to support? For some experiments rather than provide a sedml file, the authors provide python files that are meant to be run as follows:

[PathToOpenCOR]/pythonshell fig3-new.py

This would require a change to simulator specs. We would need to encode the types of valid simulation experiment files. We would lose the structured log output unless we can treat each python file as a task.

source:

https://models.physiomeproject.org/e/7d5

[jonrkarr]

I suggest we

• Publish these projects
• Include the Python scripts in the COMBINE archives we create as supplementary files
• Not try to execute these Python scripts
• Encourage Physiome users to make more use of SED-ML

Rationale:

• It looks like this example could be done with CellML and SED-ML. The need for this appears to stem from OpenCOR's limited support for SED-ML, which we have addressed (external to OpenCOR).
• The Python files in this example are partially redundant with the SED-ML because they read the SED-ML.
• The Python scripts contain both simulations and models.
• I'm not sure the Python scripts save anything