code/examples/quickstart.py

#!/usr/bin/env python3
"""PolyCAT Dataset — Quickstart Example



Demonstrates basic data loading and visualization for the PolyCAT dataset.



Display setup: 27" 4K monitor (3840x2160) at 70 cm viewing distance (~78.5 px/deg).



Requirements:

    pip install numpy pandas matplotlib



Usage:

    python code/examples/quickstart.py

"""

import csv
from collections import defaultdict
from pathlib import Path

import numpy as np

PROJECT_ROOT = Path(__file__).resolve().parents[2]


def load_csv(path):
    """Load a CSV file as a list of dicts."""
    with open(path, "r") as f:
        return list(csv.DictReader(f))


def example_1_load_metadata():
    """Example 1: Load and explore dataset metadata."""
    print("=" * 60)
    print("Example 1: Dataset Metadata")
    print("=" * 60)

    # Load participants
    participants = load_csv(PROJECT_ROOT / "data" / "metadata" / "participants.csv")
    print(f"\nParticipants: {len(participants)}")
    for p in participants[:3]:
        print(f"  {p['participant_id']}: "
              f"{p.get('total_trials_A', '?')} trials (Part A), "
              f"{p.get('total_trials_B', '?')} trials (Part B)")
    print(f"  ... and {len(participants) - 3} more")

    # Load trials
    trials = load_csv(PROJECT_ROOT / "data" / "metadata" / "trials.csv")
    print(f"\nTotal trials: {len(trials)}")

    # Count by category
    categories = defaultdict(int)
    for t in trials:
        cat = t.get("category", "unknown")
        categories[cat] += 1
    print("  By category:")
    for cat, count in sorted(categories.items()):
        print(f"    {cat}: {count}")

    return participants, trials


def example_2_load_fixations():
    """Example 2: Load and summarize fixation data."""
    print("\n" + "=" * 60)
    print("Example 2: Fixation Data")
    print("=" * 60)

    fixations = load_csv(PROJECT_ROOT / "data" / "fixations" / "fixations_all.csv")
    print(f"\nTotal fixations: {len(fixations):,}")

    # Group by participant
    by_pid = defaultdict(list)
    for f in fixations:
        by_pid[f["participant_id"]].append(f)

    print(f"Participants: {len(by_pid)}")

    # Summary per participant
    print("\nPer-participant fixation counts:")
    for pid in sorted(by_pid.keys())[:5]:
        fix_list = by_pid[pid]
        durations = [float(f["duration_ms"]) for f in fix_list]
        print(f"  {pid}: {len(fix_list):,} fixations, "
              f"mean duration = {np.mean(durations):.0f} ms")
    print(f"  ... and {len(by_pid) - 5} more")

    return fixations


def example_3_single_trial_scanpath():
    """Example 3: Load a single trial's scanpath."""
    print("\n" + "=" * 60)
    print("Example 3: Single Trial Scanpath")
    print("=" * 60)

    # Find a participant's scanpath directory
    scanpath_dir = PROJECT_ROOT / "data" / "scanpaths"
    pid_dirs = sorted(d for d in scanpath_dir.iterdir() if d.is_dir())

    if not pid_dirs:
        print("No scanpath files found. Run export_scanpaths.py first.")
        return

    # Pick first participant, first trial
    pid_dir = pid_dirs[0]
    scanpath_files = sorted(pid_dir.glob("*.csv"))
    if not scanpath_files:
        print(f"No scanpath files in {pid_dir}")
        return

    scanpath_file = scanpath_files[0]
    scanpath = load_csv(scanpath_file)

    print(f"\nParticipant: {pid_dir.name}")
    print(f"Trial: {scanpath_file.stem}")
    print(f"Fixations in scanpath: {len(scanpath)}")

    # Display note about coordinate interpretation
    print(f"\nNote: Coordinates are in pixels on a 3840x2160 display at 70 cm.")
    print(f"      At ~78.5 px/deg, a 100 px distance = ~1.27 degrees of visual angle.")

    print("\nScanpath (right eye):")
    print(f"  {'#':>3} {'X (px)':>8} {'Y (px)':>8} {'Duration':>10} {'Eye':>4}")
    print(f"  {'-'*3:>3} {'-'*8:>8} {'-'*8:>8} {'-'*10:>10} {'-'*4:>4}")
    for fix in scanpath[:10]:
        print(f"  {fix['fixation_index']:>3} "
              f"{float(fix['x_px']):>8.1f} "
              f"{float(fix['y_px']):>8.1f} "
              f"{float(fix['duration_ms']):>8.1f} ms "
              f"{fix['eye']:>4}")
    if len(scanpath) > 10:
        print(f"  ... and {len(scanpath) - 10} more fixations")


def example_4_fixation_heatmap():
    """Example 4: Create a simple fixation heatmap (text-based)."""
    print("\n" + "=" * 60)
    print("Example 4: Fixation Spatial Distribution")
    print("=" * 60)

    # Load fixations for one participant
    fixations = load_csv(PROJECT_ROOT / "data" / "fixations" / "fixations_all.csv")

    # Get first participant's right-eye fixations
    first_pid = fixations[0]["participant_id"]
    pid_fix = [(float(f["x_px"]), float(f["y_px"]))
               for f in fixations
               if f["participant_id"] == first_pid and f["eye"] == "R"]

    print(f"\n{first_pid}: {len(pid_fix)} right-eye fixations")

    # Compute spatial statistics
    xs = np.array([p[0] for p in pid_fix])
    ys = np.array([p[1] for p in pid_fix])

    ppd = 78.5  # px/deg for 27" 4K at 70 cm
    cx, cy = 3840 / 2, 2160 / 2  # screen center

    # Distance from center in degrees
    dist_deg = np.sqrt(((xs - cx) / ppd)**2 + ((ys - cy) / ppd)**2)

    print(f"\nSpatial distribution:")
    print(f"  X range: {xs.min():.0f} - {xs.max():.0f} px "
          f"({(xs.min() - cx) / ppd:.1f} to {(xs.max() - cx) / ppd:.1f} deg)")
    print(f"  Y range: {ys.min():.0f} - {ys.max():.0f} px "
          f"({(ys.min() - cy) / ppd:.1f} to {(ys.max() - cy) / ppd:.1f} deg)")
    print(f"  Mean eccentricity: {dist_deg.mean():.1f} deg")
    print(f"  Median eccentricity: {np.median(dist_deg):.1f} deg")

    # Histogram of eccentricity
    bins = [0, 2, 4, 6, 8, 10, 15, 20, 25]
    counts, _ = np.histogram(dist_deg, bins=bins)
    print(f"\n  Eccentricity distribution:")
    for i in range(len(counts)):
        bar = "#" * int(counts[i] / max(counts) * 30)
        print(f"    {bins[i]:>4}-{bins[i+1]:>2} deg: {counts[i]:>5} {bar}")


def example_5_polygon_info():
    """Example 5: Explore polygon geometry."""
    print("\n" + "=" * 60)
    print("Example 5: Polygon Apertures")
    print("=" * 60)

    poly_csv = PROJECT_ROOT / "data" / "manifests" / "polygon_geometry.csv"
    if not poly_csv.exists():
        print(f"  Polygon geometry file not found at {poly_csv}")
        return

    polygons = load_csv(poly_csv)
    print(f"\n{len(polygons)} unique polygon shapes")
    print(f"\nFirst 5 polygons:")
    for p in polygons[:5]:
        pid = p.get("polygon_id", "?")
        cx = p.get("center_x", "?")
        cy = p.get("center_y", "?")
        print(f"  {pid}: center = ({cx}, {cy})")

    # Check for polygon JSON definitions
    poly_dir = PROJECT_ROOT / "data" / "stimuli" / "polygons"
    if poly_dir.exists():
        json_files = list(poly_dir.glob("*.json"))
        print(f"\n{len(json_files)} polygon JSON vertex definition files")
        if json_files:
            import json
            with open(json_files[0]) as f:
                sample = json.load(f)
            print(f"  Sample ({json_files[0].name}):")
            if isinstance(sample, list):
                print(f"    {len(sample)} vertices")
                if sample:
                    print(f"    First vertex: {sample[0]}")
            elif isinstance(sample, dict):
                print(f"    Keys: {list(sample.keys())}")


def main():
    print("PolyCAT Dataset — Quickstart Examples")
    print("Display: 27\" 4K (3840x2160) at 70 cm (~78.5 px/deg)")
    print()

    example_1_load_metadata()
    example_2_load_fixations()
    example_3_single_trial_scanpath()
    example_4_fixation_heatmap()
    example_5_polygon_info()

    print("\n" + "=" * 60)
    print("Quickstart complete!")
    print("=" * 60)
    print("\nFor visualization examples, see quickstart.ipynb")


if __name__ == "__main__":
    main()