jonescompneurolab
GUI: `PSD Layers` plot does not plot PSD of layers
In the original tutorials ( https://jonescompneurolab.github.io/hnn-tutorials/gamma/gamma ), we use the equivalent of the GUI's PSD Layers (3x1) Figure Layout template to illustrate some scientific points about the difference between the layers; see Figure 10 in that link.
However, in the current GUI, creating a PSD Layers (3x1) Figure shows layer-specific dipoles, NOT layer-specific PSDs, see below:
I have been told that this is a known issue, and was expected to be implemented before the next official release of the Tutorials, but I couldn't find any pre-existing Issues/PRs that describe it.
Yes definitely! This will require using the plot_psd() function with the GUI code: https://jonescompneurolab.github.io/hnn-core/stable/generated/hnn_core.viz.plot_psd.html#hnn_core.viz.plot_psd
I'd start by trying to make the desired plot on it's own (just a regular simulation in a jupyter notebook). Additionally you should install the GUI so that you can try reproducing the issue above.
Once you're ready to start modifying the code, the best place to start will be here: https://github.com/jonescompneurolab/hnn-core/blob/1ef293bdd26bceae564d3c8ed4dda075d5fc2424/hnn_core/gui/_viz_manager.py#L133
and here: https://github.com/jonescompneurolab/hnn-core/blob/1ef293bdd26bceae564d3c8ed4dda075d5fc2424/hnn_core/gui/_viz_manager.py#L22
All of the plot_types have a corresponding line of code lower down which calls the appropriate plotting function. You will need to add two new plot_type which exclusively plot 1) layer2/3 PSD, 2) layer 5 PSD. The current PSD can be left as is since it plots the combination of the layers.
Don't hesitate to tag us if you have any questions!
Hey @ntolley I'm interested in working on this project. I have gone through the tutorials, and for this particular issue, I am ready with the code changes. However, I am unable to run the GUI. Do you have any documentation for testing the GUI?
Additionally, I faced some issues while installing dependencies. I will track them down and find the necessary resolutions.
Hi @Praneetha-15. This has been removed as a 'good first issue' as we already have an open PR that solves this. Please select another one to get started!