208 lines
7.5 KiB
Python
208 lines
7.5 KiB
Python
|
#!/usr/bin/env python3
|
||
|
|
"""
|
||
|
|
Turing Pattern Analysis - Mine existing data for fractal echo signatures
|
||
|
|
FFT of snapshots, sweep correlation analysis, scale invariance check
|
||
|
|
"""
|
||
|
|
import json
|
||
|
|
import numpy as np
|
||
|
|
from PIL import Image
|
||
|
|
import os
|
||
|
|
import glob
|
||
|
|
|
||
|
|
SWEEP_PATH = "/mnt/d/Resonance_Engine/beast-build/sweep_results.csv"
|
||
|
|
SNAPSHOT_DIRS = [
|
||
|
|
"/mnt/d/Resonance_Engine/beast-build/cymatics_sweep",
|
||
|
|
"/mnt/d/Resonance_Engine/beast-build/chladni_sweep",
|
||
|
|
"/mnt/d/Resonance_Engine/sri_yantra_output",
|
||
|
|
"/mnt/d/Resonance_Engine/cymatics_output",
|
||
|
|
"/mnt/d/Resonance_Engine/turing_test"
|
||
|
|
]
|
||
|
|
|
||
|
|
def analyze_snapshot_fft(image_path):
|
||
|
|
"""Analyze spatial frequency content of snapshot."""
|
||
|
|
try:
|
||
|
|
img = Image.open(image_path).convert('L') # Grayscale
|
||
|
|
arr = np.array(img, dtype=np.float32)
|
||
|
|
|
||
|
|
# 2D FFT
|
||
|
|
fft = np.fft.fft2(arr)
|
||
|
|
fft_shift = np.fft.fftshift(fft)
|
||
|
|
magnitude = np.abs(fft_shift)
|
||
|
|
|
||
|
|
# Radial average (power spectrum)
|
||
|
|
h, w = magnitude.shape
|
||
|
|
center = (h//2, w//2)
|
||
|
|
|
||
|
|
# Create radial bins
|
||
|
|
y, x = np.ogrid[:h, :w]
|
||
|
|
r = np.sqrt((x-center[1])**2 + (y-center[0])**2).astype(int)
|
||
|
|
|
||
|
|
radial_sum = np.bincount(r.ravel(), magnitude.ravel())
|
||
|
|
radial_count = np.bincount(r.ravel())
|
||
|
|
radial_profile = radial_sum / (radial_count + 1e-10)
|
||
|
|
|
||
|
|
# Find peaks (characteristic wavelengths)
|
||
|
|
peaks = []
|
||
|
|
for i in range(2, len(radial_profile)-2):
|
||
|
|
if radial_profile[i] > radial_profile[i-1] and radial_profile[i] > radial_profile[i+1]:
|
||
|
|
if radial_profile[i] > np.mean(radial_profile) * 1.5: # Significant peak
|
||
|
|
wavelength_pixels = max(h, w) / i if i > 0 else 0
|
||
|
|
peaks.append((i, wavelength_pixels, radial_profile[i]))
|
||
|
|
|
||
|
|
return {
|
||
|
|
'filename': os.path.basename(image_path),
|
||
|
|
'size': arr.shape,
|
||
|
|
'dominant_wavelengths': peaks[:3], # Top 3 peaks
|
||
|
|
'total_power': np.sum(magnitude)
|
||
|
|
}
|
||
|
|
except Exception as e:
|
||
|
|
return {'filename': os.path.basename(image_path), 'error': str(e)}
|
||
|
|
|
||
|
|
def load_sweep_data():
|
||
|
|
"""Load parameter sweep data."""
|
||
|
|
import csv
|
||
|
|
data = []
|
||
|
|
with open(SWEEP_PATH, 'r') as f:
|
||
|
|
reader = csv.DictReader(f)
|
||
|
|
for row in reader:
|
||
|
|
try:
|
||
|
|
# Skip header rows that got duplicated
|
||
|
|
if row['value'] == 'value':
|
||
|
|
continue
|
||
|
|
data.append({
|
||
|
|
'parameter': row['parameter'],
|
||
|
|
'value': float(row['value']),
|
||
|
|
'coh_mean': float(row['coh_mean']),
|
||
|
|
'asym_mean': float(row['asym_mean']),
|
||
|
|
'vort_mean': float(row['vort_mean']),
|
||
|
|
'vel_var_mean': float(row['vel_var_mean'])
|
||
|
|
})
|
||
|
|
except (ValueError, KeyError):
|
||
|
|
continue
|
||
|
|
return data
|
||
|
|
|
||
|
|
def find_turing_candidates(sweep_data):
|
||
|
|
"""Find sweep parameters that might produce Turing patterns."""
|
||
|
|
# Turing patterns typically have:
|
||
|
|
# - Intermediate coherence (not fully ordered, not chaotic)
|
||
|
|
# - Intermediate asymmetry (broken symmetry but stable)
|
||
|
|
# - Higher vorticity (rotational structures)
|
||
|
|
|
||
|
|
candidates = []
|
||
|
|
for d in sweep_data:
|
||
|
|
# Turing "sweet spot": coherence 0.72-0.74, asymmetry 12.5-13.5
|
||
|
|
if 0.72 <= d['coh_mean'] <= 0.74 and 12.5 <= d['asym_mean'] <= 13.5:
|
||
|
|
if d['vort_mean'] > 0.02: # Significant rotation
|
||
|
|
candidates.append(d)
|
||
|
|
|
||
|
|
return candidates
|
||
|
|
|
||
|
|
def check_scale_invariance():
|
||
|
|
"""Check if patterns are self-similar across scales."""
|
||
|
|
print("\n" + "="*70)
|
||
|
|
print("SCALE INVARIANCE CHECK (Fractal Echo)")
|
||
|
|
print("="*70)
|
||
|
|
|
||
|
|
# Find all snapshots
|
||
|
|
all_snapshots = []
|
||
|
|
for dir_path in SNAPSHOT_DIRS:
|
||
|
|
if os.path.exists(dir_path):
|
||
|
|
pngs = glob.glob(f"{dir_path}/*.png")
|
||
|
|
all_snapshots.extend(pngs)
|
||
|
|
|
||
|
|
print(f"\nFound {len(all_snapshots)} snapshots")
|
||
|
|
|
||
|
|
if len(all_snapshots) < 2:
|
||
|
|
print("Insufficient snapshots for comparison")
|
||
|
|
return
|
||
|
|
|
||
|
|
# Analyze FFT of each
|
||
|
|
print("\nAnalyzing spatial frequency content...")
|
||
|
|
fft_results = []
|
||
|
|
for snapshot in all_snapshots[:10]: # Limit to first 10
|
||
|
|
result = analyze_snapshot_fft(snapshot)
|
||
|
|
fft_results.append(result)
|
||
|
|
if 'dominant_wavelengths' in result and result['dominant_wavelengths']:
|
||
|
|
print(f"\n{result['filename']}:")
|
||
|
|
for peak in result['dominant_wavelengths']:
|
||
|
|
freq_bin, wavelength, power = peak
|
||
|
|
print(f" Peak at wavelength ~{wavelength:.1f} pixels (power={power:.2e})")
|
||
|
|
|
||
|
|
# Check for common wavelengths (fractal echo)
|
||
|
|
all_wavelengths = []
|
||
|
|
for r in fft_results:
|
||
|
|
if 'dominant_wavelengths' in r:
|
||
|
|
for peak in r['dominant_wavelengths']:
|
||
|
|
all_wavelengths.append(peak[1])
|
||
|
|
|
||
|
|
if all_wavelengths:
|
||
|
|
print(f"\nWavelength distribution:")
|
||
|
|
print(f" Range: {min(all_wavelengths):.1f} - {max(all_wavelengths):.1f} pixels")
|
||
|
|
|
||
|
|
# Look for power-of-2 relationships (fractal echo)
|
||
|
|
print(f"\n Checking for fractal echo (power-of-2 relationships)...")
|
||
|
|
for i, w1 in enumerate(all_wavelengths):
|
||
|
|
for w2 in all_wavelengths[i+1:]:
|
||
|
|
ratio = max(w1, w2) / min(w1, w2)
|
||
|
|
# Check if ratio is close to 2, 4, 8, etc.
|
||
|
|
for power in [2, 4, 8, 16]:
|
||
|
|
if abs(ratio - power) < 0.3:
|
||
|
|
print(f" Found: {min(w1,w2):.1f} x {power} ≈ {max(w1,w2):.1f}")
|
||
|
|
|
||
|
|
def analyze_sweep_for_turing():
|
||
|
|
"""Analyze sweep data for Turing pattern signatures."""
|
||
|
|
print("="*70)
|
||
|
|
print("SWEEP DATA ANALYSIS - TURING CANDIDATES")
|
||
|
|
print("="*70)
|
||
|
|
|
||
|
|
sweep_data = load_sweep_data()
|
||
|
|
print(f"\nLoaded {len(sweep_data)} sweep records")
|
||
|
|
|
||
|
|
# Find candidates
|
||
|
|
candidates = find_turing_candidates(sweep_data)
|
||
|
|
print(f"\nFound {len(candidates)} Turing pattern candidates:")
|
||
|
|
print(" (Coherence 0.72-0.74, Asymmetry 12.5-13.5, Vorticity > 0.02)")
|
||
|
|
|
||
|
|
for c in candidates[:10]: # Show first 10
|
||
|
|
print(f"\n {c['parameter']} = {c['value']:.4f}:")
|
||
|
|
print(f" Coherence: {c['coh_mean']:.4f}")
|
||
|
|
print(f" Asymmetry: {c['asym_mean']:.4f}")
|
||
|
|
print(f" Vorticity: {c['vort_mean']:.4f}")
|
||
|
|
print(f" Velocity variance: {c['vel_var_mean']:.6f}")
|
||
|
|
|
||
|
|
# Check khra_amp specifically
|
||
|
|
print("\n" + "="*70)
|
||
|
|
print("KHRA AMPLITUDE SWEEP - DETAILED")
|
||
|
|
print("="*70)
|
||
|
|
|
||
|
|
khra_data = [d for d in sweep_data if d['parameter'] == 'khra_amp']
|
||
|
|
khra_data.sort(key=lambda x: x['value'])
|
||
|
|
|
||
|
|
print(f"\nTesting {len(khra_data)} Khra values...")
|
||
|
|
for d in khra_data:
|
||
|
|
marker = "*** TURING CANDIDATE ***" if d in candidates else ""
|
||
|
|
print(f" Khra {d['value']:.4f}: Coh {d['coh_mean']:.4f}, Asym {d['asym_mean']:.4f} {marker}")
|
||
|
|
|
||
|
|
def main():
|
||
|
|
print("="*70)
|
||
|
|
print("TURING PATTERN ANALYSIS - FRACTAL ECHO SEARCH")
|
||
|
|
print("="*70)
|
||
|
|
|
||
|
|
# Analyze sweep data
|
||
|
|
analyze_sweep_for_turing()
|
||
|
|
|
||
|
|
# Check snapshots for scale invariance
|
||
|
|
check_scale_invariance()
|
||
|
|
|
||
|
|
print("\n" + "="*70)
|
||
|
|
print("ANALYSIS COMPLETE")
|
||
|
|
print("="*70)
|
||
|
|
print("\nKey findings:")
|
||
|
|
print(" 1. Turing candidates identified in sweep data")
|
||
|
|
print(" 2. Spatial frequency analysis of snapshots")
|
||
|
|
print(" 3. Fractal echo (scale invariance) check")
|
||
|
|
print("\nReview the candidate parameters above.")
|
||
|
|
|
||
|
|
if __name__ == '__main__':
|
||
|
|
main()
|