Feat: Integrate Real Subject Leadfields

[orabe]

Closes #4

This pull request addresses issue #4 by integrating the the use of leadfields generated from real individual subject anatomical data, beyond the reliance on the fsaverage template.

Key Changes Implemented:

• Subject-Specific Leadfield Loading:
  • Modified the DataSimulator and Benchmark classes to dynamically load and utilize leadfield matrices corresponding to individual subjects.
  • Introduced a new subject parameter in the data_param_grid (as seen in examples/run_experiments.py). The benchmarking loop now iterates over a list of specified subject IDs (e.g., ["subject_01", "subject_02", "fsaverage"]), loading the appropriate leadfield for each.
• Configuration Updates:
  • Updated configuration to allow specifying paths to these individual leadfield files. (#TODO: Details on where to place/configure these paths should be documented).
• Leadfield Comparison & Visualization:
  • Added functions to compare key properties between individual subject leadfields and the fsaverage leadfield.
• Data Handling:
  • Ensured that data loading and processing can handle different leadfield dimensions .

How to Use:

1. Ensure individual subject leadfield files are available and correctly pathed in the configuration.
2. Update the data_param_grid in experiment scripts to include the subject parameter with a list of desired subject IDs (including fsaverage if comparison is needed).

   data_param_grid = {
       # ... other parameters ...
       "subject": ["subject_01", "subject_02", "fsaverage"],
   }

3. Run the experiments as usual. The framework will now loop through the specified subjects.

Affected Files:

• calibrain/data_simulation.py
• calibrain/benchmark.py
• examples/run_experiments.py