The Week 13
Greetings,
Almost towards the end, we didn't have anything new to discuss in the last meeting, except the identifier (id) of COBRA model components. The 'id' of an SBML component is an important attribute and should not be manipulated. COBRApy, however, allows the user to pass a replacement function to manipulate the component ids. It also asks the user to pass a reverse replacement function to modify the ids while writing the model back. We discussed this thing and both of the mentors were having an opinion that ids should not be manipulated at all, it can create problems. Currently, the COBRApy writes the same manipulated ids in the JSON format, and if we apply the defined id pattern properties of SBML to these manipulated ids, these ids may get failed. So at the time of reading/writing the model from JSON, I have used the default replacement functions to fix the ids.
Another thing which we discussed was the things which COBRApy communities have listed to port the code from python 2 to python 3. They majorly involved removing "future imports", using "f strings" for logging, using type annotations, etc. Initially, I was using a python tool called "2to3" for porting from Python 2 to 3, but it was not enough. These changes will have to be done manually.
I shall also give a final touch on the documentation of the new implemented features and then we shall send the final pull request.
Thanks for reading. See you next week.
Comments
Post a Comment