## Yateslab code organization

### Organizing neuroviskit vs. projects
- Datasets / Data management
- each dataset should have a repo that contains the preprocessing code and data serving code
- dataset repo should have plotting code as well
- should fixational-transients be pip installable? Yes
- Maybe make the mitchell datasets repo pip installable instead. offload lots of things to that repo and to Neuroviskit
- NeuroVisKit
- Plotting code should be organized nicely
- Universal grid plotting function that takes in tensor/array X shape "row x col x h x w"
- Universal GIF generator takes in X and axis names "c1txy", accepts 1 in places that should be unsqueezed.
- Universal spatiotemporal weight plotting function that takes X and axis names "c1txy" to plot weights using jakes code
- Add a bunch of comments and make code readable
- TODOS
- Transfer NeuroVisKit transients code to fixational transients.

### todo: within a week, else share code snippets
- [x] Dekel: dataset is pip installable: Data-Mitchell-FreeViewing
- [x] Dekel: test the fixational-transients scripts still run with new organization
- [ ] Jake/Dekel: organize plotting code into Neuroviskit as specified above

### Code to move to neuroviskit
`get_sta` from `fixation.rf`
`plot_weights` and dependencies from `fixation.rf`

Repo naming convention:
Data-<Labname>-<Task/Stimulus>

Data-Rieke-DovesMEA
Data-Huk-Images