parameter_heatmaps.py

#!/usr/bin/env python3
"""
Plot 6 pairwise 2D heatmaps (1v1) for the 4 process parameters across all
simulations, and save the parameter list as JSON for later merging.

Outputs (written next to this script):
  parameter_list.json            — list of dicts, one per simulation
  parameter_heatmaps.png         — 2×3 grid of 2D histograms
"""
from __future__ import annotations

import json
from itertools import combinations
from pathlib import Path

import matplotlib.pyplot as plt
import numpy as np

# ── config ────────────────────────────────────────────────────────────────────

HERE      = Path(__file__).parent
DATA_DIR  = HERE / "final_data_processed"
OUT_JSON  = HERE / "parameter_list.json"
OUT_PNG   = HERE / "parameter_heatmaps.png"
BINS      = 15

PARAMS = [
    ("laser_power",          "Laser Power (W)"),
    ("scan_speed_x",         "Scan Speed (m/s)"),
    ("laser_spot_size",      "Spot Size (m)"),
    ("substrate_temperature","Substrate Temp (K)"),
]
PARAM_KEYS   = [k for k, _ in PARAMS]
PARAM_LABELS = {k: lbl for k, lbl in PARAMS}

# ── load ──────────────────────────────────────────────────────────────────────

records = []
for sim_dir in sorted(DATA_DIR.iterdir()):
    pjson = sim_dir / "parameters.json"
    if not pjson.exists():
        continue
    raw = json.loads(pjson.read_text())
    try:
        rec = {
            "sim": sim_dir.name,
            "laser_power":           float(raw["laser_power"]["value"]),
            "scan_speed_x":          float(raw["scan_speed_x"]["value"]),
            "laser_spot_size":       float(raw["laser_spot_size"]["value"]),
            "substrate_temperature": float(raw["substrate_temperature"]["value"]),
        }
    except (KeyError, ValueError) as e:
        print(f"  SKIP {sim_dir.name}: {e}")
        continue
    records.append(rec)

print(f"Loaded {len(records)} simulations")

# ── save JSON list ────────────────────────────────────────────────────────────

OUT_JSON.write_text(json.dumps(records, indent=2))
print(f"Saved parameter list → {OUT_JSON}")

# ── parameter domains ─────────────────────────────────────────────────────────

values_raw = {k: np.array([r[k] for r in records]) for k in PARAM_KEYS}

print("\nParameter domains:")
print(f"  {'Parameter':<25} {'Min':>14} {'Max':>14}")
print("  " + "-" * 55)
for k, lbl in PARAMS:
    vmin, vmax = values_raw[k].min(), values_raw[k].max()
    print(f"  {lbl:<25} {vmin:>14.6g} {vmax:>14.6g}")

# Normalize each parameter to [0, 1] over its observed domain
def normalize(v: np.ndarray) -> np.ndarray:
    lo, hi = v.min(), v.max()
    return (v - lo) / (hi - lo) if hi > lo else np.zeros_like(v)

values = {k: normalize(values_raw[k]) for k in PARAM_KEYS}

# ── heatmaps ──────────────────────────────────────────────────────────────────

pairs = list(combinations(PARAM_KEYS, 2))   # 6 pairs
assert len(pairs) == 6

fig, axes = plt.subplots(2, 3, figsize=(14, 9))
axes = axes.flatten()

for ax, (kx, ky) in zip(axes, pairs):
    x, y = values[kx], values[ky]
    h, xe, ye = np.histogram2d(x, y, bins=BINS, range=[[0, 1], [0, 1]])
    im = ax.imshow(
        h.T, origin="lower", aspect="auto", cmap="YlOrRd",
        extent=[0, 1, 0, 1],
        interpolation="nearest",
        vmin=0,
    )
    fig.colorbar(im, ax=ax, shrink=0.85, label="# simulations")
    ax.set_xlabel(f"{PARAM_LABELS[kx]} [norm]", fontsize=8)
    ax.set_ylabel(f"{PARAM_LABELS[ky]} [norm]", fontsize=8)
    ax.tick_params(labelsize=7)
    ax.set_xlim(0, 1)
    ax.set_ylim(0, 1)

plt.suptitle(
    f"Pairwise parameter coverage — normalized domains  (N={len(records)} simulations)",
    fontsize=11, y=1.01,
)
plt.tight_layout()
plt.savefig(OUT_PNG, dpi=150, bbox_inches="tight")
print(f"\nSaved heatmaps → {OUT_PNG}")
plt.show()