Add CRCNS hc-6 MATLAB ingestion starter notebook with Neuropy-compatible conversion flow

[Copilot]

This adds a lightweight starter notebook for loading locally unpacked CRCNS hc-6 .mat files and mapping core contents into Neuropy-compatible structures. The notebook is path-driven and intentionally easy to adapt when hc-6 file naming/layout differs across local copies.

• What this PR adds

  • New notebook: EXTERNAL/TESTING/testing_notebooks/CRCNS_hc6_mat_to_neuropy_starter.ipynb
  • Focused execution flow: configure dataset root → discover .mat files → inspect keys → convert data

• MAT loading + structure inspection

  • Uses repo-style import_mat_file when available (pyphoplacecellanalysis / neuropy), with scipy.io.loadmat fallback
  • Enumerates .mat files from a user-editable base_dir
  • Prints top-level keys from a sample file to quickly verify schema

• Neuropy-compatible conversion path

  • Position conversion to Position.from_separate_arrays(...) when possible (fallback to dict)
  • Spike conversion to tabular form and grouped Neurons(...) object construction
  • Epoch conversion to DataFrame and Epoch(...) object construction
  • Optional assembly into a DataSession container when neuropy session classes are importable

• Adaptability for hc-6 variants

  • File role discovery via editable pattern matching (position, spikes, epochs)
  • Defensive field extraction helpers for common MATLAB struct encodings (dtype.names / object fields)

base_dir = Path('/path/to/local/crcns/hc-6')
mat_files = sorted(base_dir.rglob('*.mat'))

position_file = _pick_file(['*position*.mat', '*pos*.mat', '*vt*.mat'])
spikes_file = _pick_file(['*spike*.mat', '*spikes*.mat', '*cell*.mat'])
epochs_file = _pick_file(['*epoch*.mat', '*behavior*.mat', '*states*.mat'])