Alan/python lib

[A14N77]

No description provided.

[jamesaboyle]

This is looking great! I'm sure this will provide a lot of use to the average user :)

Some initial comments, as we discussed in person

• The plots and color schemes are looking good.
• I really like the raster plots and the heatmap way of visualising potentials. I honestly thing that this should be the default way of showing potentials as its consistent for both spikes and potentials.
• The data conversion for taking SANA-FE output to pandas dataframes also looks great.

As mentioned, there are two paths for SANA-FE to generate traces: either as CSV files or as raw data. It would be good to support both transparently. Our API should be able to detect which is being passed to it.

• I don't think we need a separate data class for different trace types in Python. Probably simplest to use Python internal data structures (dicts and lists) if possible and limit the complexity, I think
• We should be able to take whatever sanafe.sim() spits out as its results and plot spikes, plot potentials, convert to dataframes etc. I think the simplest path might be the most useful. But probably what you've already got - if you provide the csv filename, then its good if your function can read that csv file and plot it (basically what you've got already is great). Minimising the steps for the user is good, as is minimising the numbers of classes and custom objects/added complexity :)
• Hopefully this just means a little refactoring and shuffling of Python code

At a high level it looks nearly good to go. Feel free to work on this and your chip visualizer (with NoC info) in parallel! The chip visualization could be included or separate, whatever you feel is best.