canns-team/data-analysis-datasets

CANNS Analysis Datasets

This repository contains example datasets for the CANNS (Continuous Attractor Neural Networks) data analysis package.

Datasets

ROI_data.txt (703 KB)

• —Description: 1D CANN ROI data for bump analysis
• —Format: Text file with neural activity measurements
• —Usage: 1D CANN analysis, MCMC bump fitting
• —Example: Used in 1D CANN analysis tutorials

grid_1.npz (8.7 MB)

• —Description: Grid cell spike data with position information
• —Format: NumPy archive containing spike times, positions
• —Keys: spike (spike times), t (time), x, y (positions)
• —Usage: 2D CANN analysis, topological data analysis, circular coordinate decoding
• —Example: Primary dataset for 2D CANN tutorials

grid_2.npz (4.5 MB)

• —Description: Second grid cell dataset for comparison studies
• —Format: NumPy archive with spike and position data
• —Usage: Comparative analysis, validation studies

LeftRightdata_of(604 MB)

pretty_name: "ASA-format MEC grid-cell dataset (Open Field only)" tags:

• —neuroscience
• —grid-cells
• —MEC
• —open-field
• —continuous-attractor-neural-network
• —topological-data-analysis
• —ASA-format

Dataset Summary

This dataset contains ASA-format conversions of MEC recordings restricted to Open Field (OF) sessions only. Each session is provided as a NumPy .npz file (full session + optional module subsets) with a lightweight JSON manifest for indexing.

Original Source

• —EBRAINS dataset instance: https://search.kg.ebrains.eu/instances/4080b78d-edc5-4ae4-8144-7f6de79930ea

Files Overview

*ASAmecfullcm.npz

• —Description: ASA-format Open Field (OF) MEC session data (all units included)
• —Format: NumPy archive (.npz) following the ASA schema
• —Keys: spike (neural spikes/activity), t (time), x, y (positions), meta (session metadata)
• —Usage: Full-session analysis with ASA (CANN / TDA / decoding), baseline for module comparisons
• —Example: Primary input file to run end-to-end OF analysis in the ASA pipeline

*ASAmecgridModuleXXnN_cm.npz

• —Description: ASA-format subset files containing grid cells from a specific module (Module XX, n=N cells) for the same OF session
• —Format: NumPy archive (.npz), same ASA schema as the full file, but restricted to one module’s grid cells
• —Keys: spike, t, x, y, meta
• —Usage: Faster experiments, module-wise topology/decoding analysis, comparing manifolds across modules
• —Example: Run TDA on Module 01 vs Module 02 to compare torus quality and decoding stability

*ASAmanifest.json

• —Description: Per-session manifest summarizing the source .mat, generated ASA outputs, unit counts, and module splits
• —Format: JSON metadata index
• —Usage: Batch processing, programmatic iteration over sessions, auto-generating documentation without loading large .npz
• —Example: Parse manifests to list all sessions and locate their corresponding full/module .npz files

Reference

Please refer to the EBRAINS instance above for the original dataset description and citation requirements.

Usage

Install the CANNS package and use the datasets module:

from canns import datasets
from canns.analyzer import data_analysis

# Automatic dataset download and setup
datasets.quick_setup()

# Load specific datasets
roi_data = datasets.load_roi_data()
grid_data = datasets.load_grid_data("grid_1")

# Use with analysis tools
analyzer = data_analysis.CANNDataAnalyzer()
spikes, x, y, t = analyzer.load_spike_data(datasets.get_dataset_path("grid_1"))

Examples

See the CANNS examples for complete tutorials:

• —data_analysis_demo.py: Command-line demo
• —cann_data_analysis_tutorial.ipynb: Jupyter notebook tutorial

Citation

These datasets are derived from the CANN-data-analysis repository. Please cite:

@software{cann_data_analysis,
  title = {CANN Data Analysis},
  url = {https://github.com/Airs702/CANN-data-analysis},
  author = {Airs702},
  year = {2024}
}

License

Please refer to the original CANN-data-analysis repository for license information.

Generated automatically for the CANNS package datasets.