llm-analysis Cline Rules

Your objective is to create an analysis that demonstrates the foundation systems neuroscience finding.

1. Use the 'search_dandi' and 'semantic_search_dandi' tools to identify appropriate datasets.

All new code for this analysis should be placed in a folder called named for the analysis you are performing (don't edit code in other directories)

2. Set up data streaming using LINDI and explore the structure of the data files. Identify how to read in the appropriate data streams.
3. Develop a metric for this tuning property and evaluate this across cells in the file.
    a. Run the code regularly and check the results to make sure you are on the right track. Adjust as necessary before moving forward.
    b. Make the code modular, with smaller functions across multiple files. Separate data reading, processing, and visualization.
4. Produce a figure that demonstrates the encoding feature for an individual neuron using matplotlib. This should be the neuron with the highest score selectivity score.
5. Also produce a figure that shows the distribution of this metric across all the neurons.
6. Keep in mind that it may make sense to go back to step 1 at any point if the dataset you identified isn't working.
7. Establish that the code is working by running it and examining the outputs before proceeding to the next step.
8. Once you have an analysis that demonstrates the phenomenon, assemble the code into an educational illustrative notebook that walks the user through the calculation and could act as a chapter of a textbook. Once they are satisfied, you can convert to notebook.ipynb using jupytext --to notebook notebook.py