Upload 14 files

analysis/analyze_lcz_distribution.py

@@ -0,0 +1,122 @@
+#!/usr/bin/env python3
+"""LCZ Class Distribution Analysis - Bar Chart"""
+
+import numpy as np
+import rasterio
+from rasterio.windows import Window
+import matplotlib.pyplot as plt
+from joblib import Parallel, delayed
+from pathlib import Path
+
+# Configuration
+LCZ_FILE = Path("/workspace/storage/lst-earthformer/CONUS_LCZ.tif")
+OUTPUT_DIR = Path("/workspace/storage/lst-earthformer")
+N_JOBS = 124
+MAX_CLASS = 19
+
+LCZ_LABELS = {
+    1: "Compact high-rise",
+    2: "Compact mid-rise",
+    3: "Compact low-rise",
+    4: "Open high-rise",
+    5: "Open mid-rise",
+    6: "Open low-rise",
+    8: "Large low-rise",
+    10: "Heavy industry",
+    11: "Dense trees",
+    12: "Scattered trees",
+    13: "Bush, scrub",
+    14: "Low plants",
+    15: "Bare rock",
+    16: "Bare soil",
+    17: "Water",
+    18: "Custom/Unknown",
+}
+
+def process_chunk(lcz_file, y_start, chunk_height, width):
+    """Process a horizontal strip of the raster."""
+    try:
+        with rasterio.open(lcz_file) as src:
+            window = Window(0, y_start, width, chunk_height)
+            data = src.read(1, window=window).flatten()
+            data = data[data != 0]  # Exclude class 0 (no data)
+            counts = np.bincount(data.astype(np.int64), minlength=MAX_CLASS)
+            return counts
+    except Exception as e:
+        print(f"Error processing chunk at row {y_start}: {e}")
+        return np.zeros(MAX_CLASS, dtype=np.int64)
+
+def main():
+    print("Opening LCZ raster...")
+    with rasterio.open(LCZ_FILE) as src:
+        height, width = src.height, src.width
+    print(f"Dimensions: {width:,} x {height:,} pixels ({width * height:,} total)")
+
+    # Divide into chunks
+    chunk_size = height // N_JOBS
+    chunks = []
+    for i in range(N_JOBS):
+        y_start = i * chunk_size
+        chunk_height = chunk_size if i < N_JOBS - 1 else height - y_start
+        chunks.append((y_start, chunk_height))
+
+    print(f"Processing {len(chunks)} chunks with {N_JOBS} workers...")
+    counts_list = Parallel(n_jobs=N_JOBS, prefer="processes", verbose=10)(
+        delayed(process_chunk)(str(LCZ_FILE), y, h, width) for y, h in chunks
+    )
+
+    print("Aggregating counts...")
+    total_counts = np.sum(counts_list, axis=0)
+    total_pixels = total_counts.sum()
+
+    # Filter to classes that exist (excluding class 0)
+    present_classes = [i for i in range(1, MAX_CLASS) if total_counts[i] > 0]
+    counts = [total_counts[i] for i in present_classes]
+    percentages = [c / total_pixels * 100 for c in counts]
+
+    # Dominant class
+    dominant_idx = np.argmax(counts)
+    dominant_class = present_classes[dominant_idx]
+
+    # Ranking by coverage
+    ranking = sorted(zip(present_classes, counts, percentages), key=lambda x: -x[1])
+
+    # Plot
+    print("Creating plot...")
+    plt.figure(figsize=(14, 8))
+    bars = plt.bar(range(len(present_classes)), counts, color="forestgreen", edgecolor="none")
+    plt.xticks(range(len(present_classes)), [str(c) for c in present_classes], fontsize=10)
+    plt.xlabel("LCZ Class", fontsize=12)
+    plt.ylabel("Pixel Count", fontsize=12)
+    plt.title(f"LCZ Class Distribution ({total_pixels:,} pixels, excluding class 0)", fontsize=14)
+    plt.grid(axis="y", alpha=0.3)
+
+    for bar, pct in zip(bars, percentages):
+        if pct >= 1:
+            plt.text(bar.get_x() + bar.get_width() / 2, bar.get_height(),
+                     f"{pct:.1f}%", ha="center", va="bottom", fontsize=8)
+
+    plt.tight_layout()
+    plt.savefig(OUTPUT_DIR / "lcz_class_distribution.png", dpi=150)
+    plt.close()
+    print(f"Saved: lcz_class_distribution.png")
+
+    # Statistics file
+    with open(OUTPUT_DIR / "lcz_class_stats.txt", "w") as f:
+        f.write("LCZ Class Distribution Statistics\n")
+        f.write("=" * 60 + "\n\n")
+        f.write(f"Image dimensions: {width:,} x {height:,}\n")
+        f.write(f"Total pixels (excluding class 0): {total_pixels:,}\n\n")
+        f.write(f"Dominant class: {dominant_class} ({LCZ_LABELS.get(dominant_class, 'Unknown')})\n\n")
+        f.write("Class Ranking by Area Coverage:\n")
+        f.write("-" * 60 + "\n")
+        f.write("Rank  Class       Count    Percentage  Description\n")
+        f.write("-" * 60 + "\n")
+        for rank, (cls, cnt, pct) in enumerate(ranking, 1):
+            desc = LCZ_LABELS.get(cls, "Unknown")
+            f.write(f"{rank:<6}{cls:<8}{cnt:>15,}{pct:>11.4f}%  {desc}\n")
+        f.write("\n" + "=" * 60 + "\n")
+    print(f"Saved: lcz_class_stats.txt")
+
+if __name__ == "__main__":
+    main()

analysis/analyze_lst_distribution.py

@@ -0,0 +1,115 @@
+#!/usr/bin/env python3
+"""LST Temperature Distribution Analysis - Probability Histogram"""
+
+import glob
+import numpy as np
+import rasterio
+import matplotlib.pyplot as plt
+from joblib import Parallel, delayed
+from pathlib import Path
+
+# Configuration
+LST_DIR = Path("/workspace/storage/lst-earthformer/Data/ML/Cities_Tiles")
+OUTPUT_DIR = Path("/workspace/storage/lst-earthformer")
+N_JOBS = 124
+TEMP_MIN, TEMP_MAX = -189, 211
+BINS = range(TEMP_MIN, TEMP_MAX + 2)
+
+def process_file(filepath):
+    """Process single TIF file, return histogram counts."""
+    try:
+        with rasterio.open(filepath) as src:
+            data = src.read(1).flatten()
+            data = data[data != 0]  # Exclude 0F (no data)
+            counts, _ = np.histogram(data, bins=BINS)
+            return counts
+    except Exception as e:
+        print(f"Error processing {filepath}: {e}")
+        return np.zeros(len(BINS) - 1, dtype=np.int64)
+
+def main():
+    print("Finding LST files...")
+    files = glob.glob(str(LST_DIR / "**" / "LST_*.tif"), recursive=True)
+    total_files = len(files)
+    print(f"Found {total_files:,} files")
+
+    if total_files == 0:
+        print("No files found. Exiting.")
+        return
+
+    print(f"Processing with {N_JOBS} workers...")
+    histograms = Parallel(n_jobs=N_JOBS, prefer="processes", verbose=10)(
+        delayed(process_file)(f) for f in files
+    )
+
+    print("Aggregating histograms...")
+    total_hist = np.sum(histograms, axis=0)
+    temperatures = np.arange(TEMP_MIN, TEMP_MAX + 1)
+    total_pixels = total_hist.sum()
+
+    # Probability distribution
+    probabilities = total_hist / total_pixels
+
+    # Statistics
+    weighted_temps = temperatures * total_hist
+    mean_temp = weighted_temps.sum() / total_pixels
+    mode_idx = np.argmax(total_hist)
+    mode_temp = temperatures[mode_idx]
+
+    # Cumulative for percentiles
+    cumsum = np.cumsum(total_hist)
+    percentiles = {}
+    for p in [10, 25, 50, 75, 90]:
+        idx = np.searchsorted(cumsum, total_pixels * p / 100)
+        percentiles[p] = temperatures[min(idx, len(temperatures) - 1)]
+
+    # Find actual min/max with data
+    nonzero = np.nonzero(total_hist)[0]
+    actual_min = temperatures[nonzero[0]] if len(nonzero) > 0 else TEMP_MIN
+    actual_max = temperatures[nonzero[-1]] if len(nonzero) > 0 else TEMP_MAX
+
+    # Variance for std
+    variance = np.sum(total_hist * (temperatures - mean_temp) ** 2) / total_pixels
+    std_temp = np.sqrt(variance)
+
+    # Plot
+    print("Creating plot...")
+    plt.figure(figsize=(14, 8))
+    plt.bar(temperatures, probabilities, width=1, edgecolor="none", color="steelblue")
+    plt.xlabel("Temperature (°F)", fontsize=12)
+    plt.ylabel("Probability", fontsize=12)
+    plt.title(f"LST Temperature Distribution ({total_files:,} files, {total_pixels:,} pixels, excluding 0°F)", fontsize=14)
+    plt.grid(axis="y", alpha=0.3)
+    plt.tight_layout()
+    plt.savefig(OUTPUT_DIR / "lst_temperature_distribution.png", dpi=150)
+    plt.close()
+    print(f"Saved: lst_temperature_distribution.png")
+
+    # Statistics file
+    with open(OUTPUT_DIR / "lst_temperature_stats.txt", "w") as f:
+        f.write("LST Temperature Distribution Statistics\n")
+        f.write("=" * 50 + "\n\n")
+        f.write(f"Total files processed: {total_files:,}\n")
+        f.write(f"Total pixels (excluding 0°F): {total_pixels:,}\n\n")
+        f.write(f"Min temperature: {actual_min}°F\n")
+        f.write(f"Max temperature: {actual_max}°F\n")
+        f.write(f"Mean temperature: {mean_temp:.2f}°F\n")
+        f.write(f"Median temperature: {percentiles[50]}°F\n")
+        f.write(f"Mode temperature: {mode_temp}°F\n")
+        f.write(f"Std deviation: {std_temp:.2f}°F\n\n")
+        f.write("Percentiles:\n")
+        for p, t in percentiles.items():
+            f.write(f"  {p}th: {t}°F\n")
+        f.write("\n" + "=" * 50 + "\n")
+        f.write("Full Histogram Data\n")
+        f.write("Temp(°F)\tCount\t\tProbability\n")
+        f.write("-" * 50 + "\n")
+        for i, t in enumerate(temperatures):
+            c = total_hist[i]
+            p = probabilities[i]
+            if c > 0:
+                f.write(f"{t}\t\t{c:,}\t\t{p:.8f}\n")
+    print(f"Saved: lst_temperature_stats.txt")
+
+if __name__ == "__main__":
+    main()

analysis/analyze_lst_per_lcz.py

@@ -0,0 +1,253 @@
+#!/usr/bin/env python3
+"""LST Distribution per LCZ Class - Cross-references LST tiles with CONUS LCZ raster"""
+
+import glob
+import numpy as np
+import rasterio
+from rasterio.warp import transform_bounds, Resampling
+from rasterio.transform import from_bounds
+from rasterio.windows import from_bounds as window_from_bounds
+import matplotlib.pyplot as plt
+from joblib import Parallel, delayed
+from pathlib import Path
+import json
+
+# Configuration
+LST_DIR = Path("/workspace/storage/lst-earthformer/Data/ML/Cities_Tiles")
+LCZ_FILE = Path("/workspace/storage/lst-earthformer/CONUS_LCZ.tif")
+OUTPUT_DIR = Path("/workspace/storage/lst-earthformer")
+N_JOBS = 64
+TEMP_MIN, TEMP_MAX = -189, 211
+NUM_BINS = TEMP_MAX - TEMP_MIN + 1
+
+LCZ_LABELS = {
+    1: "Compact high-rise", 2: "Compact mid-rise", 3: "Compact low-rise",
+    4: "Open high-rise", 5: "Open mid-rise", 6: "Open low-rise",
+    8: "Large low-rise", 10: "Heavy industry",
+    11: "Dense trees", 12: "Scattered trees", 13: "Bush, scrub",
+    14: "Low plants", 15: "Bare rock", 16: "Bare soil",
+    17: "Water", 18: "Custom/Unknown",
+}
+
+# Short labels for plots
+LCZ_SHORT = {
+    1: "Compact\nhigh", 2: "Compact\nmid", 3: "Compact\nlow",
+    4: "Open\nhigh", 5: "Open\nmid", 6: "Open\nlow",
+    8: "Large\nlow", 10: "Heavy\nindustry",
+    11: "Dense\ntrees", 12: "Scattered\ntrees", 13: "Bush\nscrub",
+    14: "Low\nplants", 15: "Bare\nrock", 16: "Bare\nsoil",
+    17: "Water", 18: "Custom",
+}
+
+MAX_LCZ = 19
+
+
+def process_file(filepath, lcz_file):
+    """Process one LST tile: read LST values, sample LCZ at same locations, return per-class histograms."""
+    try:
+        with rasterio.open(filepath) as lst_src:
+            lst_data = lst_src.read(1)
+            lst_bounds = lst_src.bounds
+            lst_crs = lst_src.crs
+            lst_h, lst_w = lst_data.shape
+
+        # Transform LST bounds to LCZ CRS (4326)
+        lcz_bounds = transform_bounds(lst_crs, 'EPSG:4326',
+                                      lst_bounds.left, lst_bounds.bottom,
+                                      lst_bounds.right, lst_bounds.top)
+
+        with rasterio.open(lcz_file) as lcz_src:
+            # Get window in LCZ raster corresponding to LST tile bounds
+            try:
+                win = window_from_bounds(*lcz_bounds, transform=lcz_src.transform)
+            except Exception:
+                return np.zeros((MAX_LCZ, NUM_BINS), dtype=np.int64)
+
+            # Read LCZ data for this window
+            lcz_data = lcz_src.read(1, window=win)
+
+            if lcz_data.size == 0:
+                return np.zeros((MAX_LCZ, NUM_BINS), dtype=np.int64)
+
+            # Resize LCZ to match LST tile (128x128) using nearest neighbor
+            from PIL import Image
+            lcz_resized = np.array(
+                Image.fromarray(lcz_data).resize((lst_w, lst_h), Image.NEAREST)
+            )
+
+        # Build per-LCZ-class histograms
+        histograms = np.zeros((MAX_LCZ, NUM_BINS), dtype=np.int64)
+        bins = np.arange(TEMP_MIN, TEMP_MAX + 2)
+
+        for lcz_class in range(1, MAX_LCZ):
+            mask = (lcz_resized == lcz_class) & (lst_data != 0)
+            if mask.any():
+                vals = lst_data[mask]
+                counts, _ = np.histogram(vals, bins=bins)
+                histograms[lcz_class] = counts
+
+        return histograms
+
+    except Exception as e:
+        return np.zeros((MAX_LCZ, NUM_BINS), dtype=np.int64)
+
+
+def main():
+    print("Finding LST files...")
+    files = glob.glob(str(LST_DIR / "**" / "LST_*.tif"), recursive=True)
+    total_files = len(files)
+    print(f"Found {total_files:,} files")
+
+    if total_files == 0:
+        print("No files found.")
+        return
+
+    print(f"Processing with {N_JOBS} workers...")
+    results = Parallel(n_jobs=N_JOBS, prefer="processes", verbose=10)(
+        delayed(process_file)(f, str(LCZ_FILE)) for f in files
+    )
+
+    print("Aggregating histograms...")
+    total_hists = np.sum(results, axis=0)  # shape: (MAX_LCZ, NUM_BINS)
+    temperatures = np.arange(TEMP_MIN, TEMP_MAX + 1)
+
+    # Find which LCZ classes have data
+    present_classes = [c for c in range(1, MAX_LCZ) if total_hists[c].sum() > 0]
+    print(f"LCZ classes with data: {present_classes}")
+
+    # Compute stats per class
+    stats = {}
+    for c in present_classes:
+        hist = total_hists[c]
+        total = hist.sum()
+        if total == 0:
+            continue
+        prob = hist / total
+        mean = np.sum(temperatures * hist) / total
+        var = np.sum(hist * (temperatures - mean) ** 2) / total
+        std = np.sqrt(var)
+        cumsum = np.cumsum(hist)
+        median_idx = np.searchsorted(cumsum, total * 0.5)
+        median = temperatures[min(median_idx, len(temperatures) - 1)]
+        nonzero = np.nonzero(hist)[0]
+        tmin = temperatures[nonzero[0]]
+        tmax = temperatures[nonzero[-1]]
+        stats[c] = {
+            'total_pixels': int(total),
+            'mean': float(mean),
+            'std': float(std),
+            'median': int(median),
+            'min': int(tmin),
+            'max': int(tmax),
+        }
+
+    # --- Plot 1: Overlay probability distributions ---
+    print("Creating overlay plot...")
+    fig, ax = plt.subplots(figsize=(14, 8))
+    cmap = plt.cm.tab20
+    for i, c in enumerate(present_classes):
+        hist = total_hists[c]
+        total = hist.sum()
+        if total == 0:
+            continue
+        prob = hist / total
+        label = f"LCZ {c}: {LCZ_LABELS.get(c, '?')}"
+        color = cmap(i / len(present_classes))
+        ax.plot(temperatures, prob, label=label, color=color, linewidth=1.2, alpha=0.8)
+
+    ax.set_xlabel("Temperature (F)", fontsize=12)
+    ax.set_ylabel("Probability", fontsize=12)
+    ax.set_title("LST Distribution by LCZ Class", fontsize=14)
+    ax.set_xlim(0, 180)
+    ax.legend(fontsize=8, ncol=2, loc='upper left')
+    ax.grid(axis='y', alpha=0.3)
+    plt.tight_layout()
+    plt.savefig(OUTPUT_DIR / "lst_per_lcz_overlay.png", dpi=150)
+    plt.close()
+    print("Saved: lst_per_lcz_overlay.png")
+
+    # --- Plot 2: Subplots per LCZ class ---
+    print("Creating subplot grid...")
+    n_classes = len(present_classes)
+    ncols = 4
+    nrows = (n_classes + ncols - 1) // ncols
+    fig, axes = plt.subplots(nrows, ncols, figsize=(16, 3.5 * nrows), sharex=True, sharey=False)
+    axes = axes.flatten()
+
+    for i, c in enumerate(present_classes):
+        ax = axes[i]
+        hist = total_hists[c]
+        total = hist.sum()
+        prob = hist / total
+        s = stats[c]
+        ax.bar(temperatures, prob, width=1, color='steelblue', edgecolor='none')
+        ax.set_title(f"LCZ {c}: {LCZ_LABELS.get(c, '?')}", fontsize=10)
+        ax.set_xlim(0, 180)
+        ax.axvline(s['mean'], color='red', linestyle='--', linewidth=0.8, label=f"Mean: {s['mean']:.1f}F")
+        ax.legend(fontsize=7, loc='upper left')
+        ax.grid(axis='y', alpha=0.3)
+        info = f"n={total:,}\nstd={s['std']:.1f}F"
+        ax.text(0.97, 0.95, info, transform=ax.transAxes, fontsize=7,
+                verticalalignment='top', horizontalalignment='right',
+                bbox=dict(boxstyle='round,pad=0.3', facecolor='wheat', alpha=0.5))
+
+    # Hide unused subplots
+    for i in range(n_classes, len(axes)):
+        axes[i].set_visible(False)
+
+    # Common labels
+    fig.text(0.5, 0.02, "Temperature (F)", ha='center', fontsize=12)
+    fig.text(0.02, 0.5, "Probability", va='center', rotation='vertical', fontsize=12)
+    fig.suptitle("LST Temperature Distribution by LCZ Class", fontsize=14, y=0.98)
+    plt.tight_layout(rect=[0.03, 0.03, 1, 0.96])
+    plt.savefig(OUTPUT_DIR / "lst_per_lcz_subplots.png", dpi=150)
+    plt.close()
+    print("Saved: lst_per_lcz_subplots.png")
+
+    # --- Plot 3: Box-whisker style summary ---
+    print("Creating summary plot...")
+    fig, ax = plt.subplots(figsize=(14, 6))
+    means = [stats[c]['mean'] for c in present_classes]
+    stds = [stats[c]['std'] for c in present_classes]
+    labels = [f"LCZ {c}" for c in present_classes]
+    x = range(len(present_classes))
+
+    bars = ax.bar(x, means, yerr=stds, capsize=4, color='steelblue', edgecolor='none', alpha=0.8)
+    ax.set_xticks(x)
+    ax.set_xticklabels([LCZ_SHORT.get(c, str(c)) for c in present_classes], fontsize=8)
+    ax.set_ylabel("Mean LST (F)", fontsize=12)
+    ax.set_title("Mean LST by LCZ Class (error bars = 1 std)", fontsize=14)
+    ax.grid(axis='y', alpha=0.3)
+
+    for i, (m, c) in enumerate(zip(means, present_classes)):
+        ax.text(i, m + stds[i] + 1, f"{m:.1f}", ha='center', fontsize=8)
+
+    plt.tight_layout()
+    plt.savefig(OUTPUT_DIR / "lst_per_lcz_means.png", dpi=150)
+    plt.close()
+    print("Saved: lst_per_lcz_means.png")
+
+    # --- Stats file ---
+    print("Writing stats...")
+    with open(OUTPUT_DIR / "lst_per_lcz_stats.txt", "w") as f:
+        f.write("LST Distribution per LCZ Class\n")
+        f.write("=" * 70 + "\n\n")
+        f.write(f"Total LST files processed: {total_files:,}\n")
+        f.write(f"LCZ classes with data: {len(present_classes)}\n\n")
+        f.write(f"{'Class':<8}{'Description':<25}{'Pixels':>15}{'Mean':>10}{'Std':>10}{'Median':>10}{'Min':>8}{'Max':>8}\n")
+        f.write("-" * 94 + "\n")
+        for c in present_classes:
+            s = stats[c]
+            desc = LCZ_LABELS.get(c, 'Unknown')
+            f.write(f"LCZ {c:<4}{desc:<25}{s['total_pixels']:>15,}{s['mean']:>10.2f}{s['std']:>10.2f}{s['median']:>10}{s['min']:>8}{s['max']:>8}\n")
+        f.write("\n")
+
+        total_all = sum(s['total_pixels'] for s in stats.values())
+        f.write(f"{'Total':<33}{total_all:>15,}\n")
+
+    print(f"Saved: lst_per_lcz_stats.txt")
+    print("Done!")
+
+
+if __name__ == "__main__":
+    main()

analysis/combined_distribution_plot.py

@@ -0,0 +1,151 @@
+#!/usr/bin/env python3
+"""Combined LCZ + LST distribution plot in academic style."""
+
+import numpy as np
+import matplotlib.pyplot as plt
+from matplotlib.ticker import FixedLocator, FuncFormatter
+from scipy.interpolate import PchipInterpolator, make_interp_spline
+
+plt.rcParams.update({
+    'font.family': 'serif',
+    'mathtext.fontset': 'dejavuserif',
+    'axes.linewidth': 0.6,
+    'xtick.major.width': 0.5,
+    'ytick.major.width': 0.5,
+    'xtick.major.size': 3,
+    'ytick.major.size': 3,
+})
+
+def human_format(x, pos):
+    if x >= 1e9:
+        return f'{x/1e9:g}B'
+    elif x >= 1e6:
+        return f'{x/1e6:g}M'
+    elif x >= 1e3:
+        return f'{x/1e3:g}K'
+    elif x >= 1:
+        return f'{x:g}'
+    return '0'
+
+# ── Parse LST stats ──────────────────────────────────────────────
+lst_temps, lst_counts = [], []
+in_data = False
+with open('/workspace/storage/lst-earthformer/lst_temperature_stats.txt') as f:
+    for line in f:
+        if line.startswith('Temp'):
+            in_data = True
+            continue
+        if line.startswith('---'):
+            continue
+        if in_data and line.strip():
+            parts = line.split()
+            lst_temps.append(int(parts[0]))
+            lst_counts.append(int(parts[1].replace(',', '')))
+
+lst_temps = np.array(lst_temps)
+lst_counts = np.array(lst_counts, dtype=np.float64)
+
+mask = (lst_temps >= -50) & (lst_temps <= 175)
+lst_temps_trim = lst_temps[mask]
+lst_counts_trim = lst_counts[mask]
+
+bin_width = 5
+bin_edges = np.arange(-50, 176, bin_width)
+bin_centers = bin_edges[:-1] + bin_width / 2
+binned_counts = np.zeros(len(bin_centers), dtype=np.float64)
+for i, (lo, hi) in enumerate(zip(bin_edges[:-1], bin_edges[1:])):
+    sel = (lst_temps_trim >= lo) & (lst_temps_trim < hi)
+    binned_counts[i] = lst_counts_trim[sel].sum()
+
+# ── Parse LCZ stats ─────────────────────────────────────────────
+lcz_classes, lcz_counts = [], []
+in_data = False
+with open('/workspace/storage/lst-earthformer/lcz_class_stats.txt') as f:
+    for line in f:
+        if line.startswith('Rank'):
+            in_data = True
+            continue
+        if line.startswith('---') or line.startswith('='):
+            if in_data and line.startswith('='):
+                in_data = False
+            continue
+        if in_data and line.strip():
+            parts = line.split()
+            lcz_classes.append(int(parts[1]))
+            lcz_counts.append(int(parts[2].replace(',', '')))
+
+order = np.argsort(lcz_classes)
+lcz_classes = np.array(lcz_classes)[order]
+lcz_counts = np.array(lcz_counts, dtype=np.float64)[order]
+total_lcz = lcz_counts.sum()
+
+# ── Figure ───────────────────────────────────────────────────────
+fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(12, 5.0))
+fig.subplots_adjust(wspace=0.22, left=0.08, right=0.99, top=0.97, bottom=0.12)
+
+bar_color = '#6BAED6'
+curve_color = '#2171B5'
+fmt = FuncFormatter(human_format)
+
+# ── Left: LCZ class distribution ────────────────────────────────
+x_pos = np.arange(len(lcz_classes))
+ax1.bar(x_pos, lcz_counts, width=0.75, color=bar_color, edgecolor='white', linewidth=0.3)
+
+# PCHIP: touches every bar top, smooth, no overshoot
+pchip = PchipInterpolator(x_pos, lcz_counts)
+x_fine = np.linspace(x_pos[0], x_pos[-1], 300)
+y_fine = np.maximum(pchip(x_fine), 0)
+ax1.plot(x_fine, y_fine, color=curve_color, linewidth=1.0)
+
+ax1.yaxis.set_major_locator(FixedLocator([0, 1.2e6, 2.4e6]))
+ax1.yaxis.set_major_formatter(fmt)
+ax1.set_ylim(0, 3e6)
+ax1.set_xticks(x_pos)
+ax1.set_xticklabels([str(c) for c in lcz_classes], fontsize=16)
+ax1.set_xlabel('LCZ Class', fontsize=16)
+ax1.set_ylabel('Pixels', fontsize=16)
+ax1.tick_params(axis='both', labelsize=16)
+ax1.set_xlim(-0.6, len(lcz_classes) - 0.4)
+
+dominant_cls = lcz_classes[np.argmax(lcz_counts)]
+dominant_pct = lcz_counts.max() / total_lcz * 100
+txt = f'Total pixels: ~{int(round(total_lcz, -6) // 1e6):.0f}M\nDominant: Class {dominant_cls} ({dominant_pct:.1f}%)'
+ax1.text(0.97, 0.97, txt, transform=ax1.transAxes, fontsize=16,
+         verticalalignment='top', horizontalalignment='right',
+         bbox=dict(boxstyle='round,pad=0.3', facecolor='white', edgecolor='0.6', linewidth=0.5))
+ax1.text(0.97, 0.78, 'LCZ Distribution', transform=ax1.transAxes, fontsize=16,
+         color='red', fontweight='bold', verticalalignment='top', horizontalalignment='right')
+
+
+# ── Right: LST temperature distribution ──────────────────────────
+ax2.bar(bin_centers, binned_counts, width=bin_width * 0.85, color=bar_color, edgecolor='white', linewidth=0.3)
+
+# PCHIP for LST too
+pchip2 = PchipInterpolator(bin_centers, binned_counts)
+x_smooth2 = np.linspace(bin_centers[0], bin_centers[-1], 500)
+y_smooth2 = np.maximum(pchip2(x_smooth2), 0)
+ax2.plot(x_smooth2, y_smooth2, color=curve_color, linewidth=1.0)
+
+ax2.yaxis.set_major_locator(FixedLocator([0, 5e8, 1e9, 1.5e9, 2e9]))
+ax2.yaxis.set_major_formatter(fmt)
+ax2.set_ylim(0, 2.2e9)
+ax2.set_xlabel('Temperature (\u00b0F)', fontsize=16)
+ax2.set_ylabel('Pixels', fontsize=16)
+ax2.tick_params(axis='both', labelsize=16)
+ax2.set_xlim(-55, 180)
+
+total_pixels_lst = int(binned_counts.sum())
+mean_t = np.average(bin_centers, weights=binned_counts)
+variance = np.average((bin_centers - mean_t)**2, weights=binned_counts)
+std_t = np.sqrt(variance)
+txt2 = f'Total pixels: ~{total_pixels_lst / 1e9:.1f}B\nMean: {mean_t:.1f}\u00b0F  Std: {std_t:.1f}\u00b0F'
+ax2.text(0.97, 0.97, txt2, transform=ax2.transAxes, fontsize=16,
+         verticalalignment='top', horizontalalignment='right',
+         bbox=dict(boxstyle='round,pad=0.3', facecolor='white', edgecolor='0.6', linewidth=0.5))
+ax2.text(0.97, 0.78, 'Temperature Distribution', transform=ax2.transAxes, fontsize=16,
+         color='red', fontweight='bold', verticalalignment='top', horizontalalignment='right')
+
+
+plt.savefig('/workspace/storage/lst-earthformer/combined_distribution.png', dpi=300, facecolor='white')
+plt.close()
+print('Saved combined_distribution.png')

analysis/out/kdd_evaluation_report.pdf

@@ -0,0 +1,118 @@
+%PDF-1.4
+%���� ReportLab Generated PDF document (opensource)
+1 0 obj
+<<
+/F1 2 0 R /F2 3 0 R /F3 4 0 R
+>>
+endobj
+2 0 obj
+<<
+/BaseFont /Helvetica /Encoding /WinAnsiEncoding /Name /F1 /Subtype /Type1 /Type /Font
+>>
+endobj
+3 0 obj
+<<
+/BaseFont /Helvetica-Bold /Encoding /WinAnsiEncoding /Name /F2 /Subtype /Type1 /Type /Font
+>>
+endobj
+4 0 obj
+<<
+/BaseFont /Helvetica-BoldOblique /Encoding /WinAnsiEncoding /Name /F3 /Subtype /Type1 /Type /Font
+>>
+endobj
+5 0 obj
+<<
+/Contents 11 0 R /MediaBox [ 0 0 612 792 ] /Parent 10 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+6 0 obj
+<<
+/Contents 12 0 R /MediaBox [ 0 0 612 792 ] /Parent 10 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+7 0 obj
+<<
+/Contents 13 0 R /MediaBox [ 0 0 612 792 ] /Parent 10 0 R /Resources <<
+/Font 1 0 R /ProcSet [ /PDF /Text /ImageB /ImageC /ImageI ]
+>> /Rotate 0 /Trans <<
+
+>> 
+  /Type /Page
+>>
+endobj
+8 0 obj
+<<
+/PageMode /UseNone /Pages 10 0 R /Type /Catalog
+>>
+endobj
+9 0 obj
+<<
+/Author (\(anonymous\)) /CreationDate (D:20260207010054+00'00') /Creator (\(unspecified\)) /Keywords () /ModDate (D:20260207010054+00'00') /Producer (ReportLab PDF Library - \(opensource\)) 
+  /Subject (\(unspecified\)) /Title (\(anonymous\)) /Trapped /False
+>>
+endobj
+10 0 obj
+<<
+/Count 3 /Kids [ 5 0 R 6 0 R 7 0 R ] /Type /Pages
+>>
+endobj
+11 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 2634
+>>
+stream
+Gb"/(92jn2'#*rfo!/kH">\QG0.-N\"$JBN6h_)?c<fdF>UmmU+1*,oCs7s4PiYtDa9Ei7@h0Rp<!b"=8ps<EIVDZ`nO&f7-jBCUK]>iB."gUcW3?CUItN[@_7'O#8rCDni%a!WE<>jg9n./0J5^0cc%BSbRM\jkBjsd##s=_m_f?3EJ"3.KktMdkELt!h<Ni$F^f$I5>Wd[2Xo#<_I9om.Yl"jOT7@@!d@fCeo@t`uF@KDn9`pa!oAA_CEc!.WT\46njl50+[f-R4/Fou"W_VLEo^l55r^W]"Qt8#r@;@s(d,rWjQd'OQ2N8I;N:7c2m)B=EG;p*`GljW!9i6\b'A2.R0UoU/.saLjir28JJ%o6$&hO"fPBbo"l9qbj#_$cDAuWbAq?NHFC;trGGP9(80Q%S$Xp!s[^ouF>N>`Ao1s`1=\L6mCY-`itfGhR6rr\n.S=_E0:Va782@"rb5saHAs8Ljr_UU:E=jopXU"j[nQ,*'N.YKa/IW"W?qPj"S"/H^&BkJ;'IC4lo\u]EC+!=7?e_uJW1mhTP4^@m`H^:T(@1ej9@(q2O%'-t$:t@S%X@fG734jn00o)uh[W."-NPhRrVb)JrY@UTuV%-A*!B%4gImA)&I*)TilPcM-MIj=ZZs^G0E"Nl9G*<^kmOKn5lo)72X)IQRN?44IHMD-O;#]hQBCagMo!/j^o?0A$Q_.i#9k3,b1KXnl"`GAc`NJnKk`1(lIA8,a_&);P6F_GGTcS8C!#?6<EXs_oN=Q'DOu::2AdJ(:'938hf_gE&e>EsfqZ9e@..piL!lst>Nk]n"G"OH`dXAGk5,#4fs/ZYtXH!0'QsK&q6@0XsYA31K(cgNaM$uIJNoj2./_88He@O_KX:4(hAffRm-_,$\Q=GkV%RW5t1iga#X39f:P;KcH@0`ieNcNoc2"Ng[YQsPiHC@\!;`&f%4U+KuUn.tKidn^Or8jnD3Tks)^VP(730ra<]=+3:#Hm16R6H)dL2?OgPLt339jFJ>c0MIJGX?d;IX-2TR6`'IT-nWhQ"tsb1$,*eO^Q>*g1`$W[6@TZ*6PT-Y@#>CP]7*J6*HU66c)Q1m9AF<P$<K'Opg/fl*.H^m?\!1m94<WkSQ.fi+[;NLNngB+pjqVd(3E<#-ljE1<UbX0U*53M/,39qAthf&ZWX=ja\N/pis:;[e'E2GhjDt_5?03,^k<Y*P?9FRdd35*<_Ak$4ZG;qX#j`e,2rbP;*0`[Diimonfc.n.lOfd4'=p)$hX<ac`Q99NtqX_@BN$,NYAa]>_.@`7/CLWl((q0'c2@g'bKHPK/KP0muGb]2^9$0I9g8(dT0P.M\V2e'`b]l)S/T^&h(^CO3-f3!YBN:MY_t/*L*<_3I$d#<iFI(,[WpC^_LW_/Bd*V=<#;g*!)s*2L(b&P:_.YM+leG&17Sme5<YWG_X1q22rZrG:[.L@/I5Fba:IkiG/bms6DaaI4u/CKQes?8p<>L(T1!=YqA72&c#WY12pebF-tdr3U=#7f7.Hh<IIYIu/pfk.Xr%d@K9S4YNt`[PGG+cf;*Yq]^NO=dVYKUcI?sQS'CYTfK*I]-m0*KjR--Y,kTK>9+bN,`=^G@7H?Qf=-nKJR_uA<WYi0@I!2K3KA7p7ZE^pnYX`L2ej3f]6e!$PK-[Lik?l4FLpm/^A"aiUV!E&SY>J'[KY1;#S1N>.hemp6P6S,F;;_6O_7#Y+S%lfr^#Up&Gaojq+Tme=_'q6(a+2ME<g!bGnV!0_C&:VpSI@QS]pCJ4T8ud3F)t&qIK&Nko;mO"fBgs&q?D3O"p3!giqU#"Yp*r94#Bq8b1B\87W@frs=\*-eFt68V<fm!7sNl&UmF#1D"bnY*&5I7"1?\\GJHEc858W9elB[_uE><E/F&.j3_d.'E$1$Hsd-G1oS(]gL(68Tf-iuFLGiQj^"B%MJe*.)0+hqHr\Z/\]9=JL77/7np'5BF9]ujco%+C]0%.:2&6&$4<>#2N1Z+P4T>A6a"o'V+-04?.ODPZM_XR_8ERTqAD@qqX>HL:;r&%]CZbg%HF)9mS11'/&12Ca<MYEm0F2t%)9H6@`>pDnXHXIifIk55q0O#-CQd7[*"e8"o6bB]WfK*D>f4fb--7JeLh]2"M^]]-L?=1u;8o"YWZ[G6+'e_XS>hP1I3*5i@%M0.`fHJmNWT-bD7`ZHD@*2`"%F:uH(\F(*B,XqJhqgj5nn@La_6(&\69$^Q8=?spb_.$A#T"%qj<9qg?P.BOP1+k[B^:C&2nN[L13Z<6O779Dkc*:etqa$?`]f&N:r_&&L[9r2>=W:<qtjl3pJj0<N4.WFn_*@esP!^o?,VBY%:L6gg*qEMkc\-7!e^$8rc'HSf)C^\/n'cn%JEe"p:0ck(]=_)b9)W-iUjNai=tE:-I@dkZH8#hN&D6GYLIlZ,!VkPNM0SPo]6upKT_sKrQTAb,;6_(3%G'gp,hXEeO\@@^!7F]pK3jGK4TB.>307c^=Zs9'D6n)K3C;:nF@nKu9sYB0s<L^J(0-T+H,B13)<"_PWVp4*/ZU'-'B3(s+f,IlZ3/$2)r]ccJi;TgfJhH>&(eW31]'Y`;KB<^3.d:<OSuWuk/V'/XRQX^T#S>6^p.K6c6jf?c*nbVS280Aq"&?2~>endstream
+endobj
+12 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 3175
+>>
+stream
+GauHN>BAQ/'n5n\^n\4nBEI+7BDSju7,)FHlnn7>1EF4q&gVHaOX`Z>j8JX6\<58EPjN]VhMiE2i8EMqE6e3?+Xi!cUAnJ7&c2/$h9H9M'KrN/M1Q!rgnRXU=9#rXerk\(8gk-(Yh..bVcdA,WgAHrjU&dT7#aI4T<<c.+s]lZON0qY0_h3e36L'-r7M"a?1LU+m+=MO0gi-M<i/qem@']q/Ll<_jlg-2Y1KB\:#/Al2geam@2bio&*Vs_o^L>OQornd)Vtb4.EW:Ppq/&L)'EY<M8ps$!g_<l.LkD$]HrP.O:RXl0;uoHc\$K'P6Om5hNGil5Xq&B5Wj$5-oFPPeZ)^cU2"[L'mn>!JSF*K.l!Va]+L%\-%;l[-)To@@!7?=E=T4K?X2TAhZ;jjj+)BIAK_@8dK^.GGQMDq*^kWh*`+LpOb'R8]CcBsSf>Bq@,[Lq.;!sn?,[aFDhRm!OS.t0!QpF1dm.%dP$A-IooR1bH/6S_Q=5mO+LO`p#AK_'(/>99qa)&UXSj$?$Fj5XS%VW9T,UYq<aCP`oboXDp!Q)CHV^FIVge/=3";'/KHkQs34VCC4?MUXEY)gB;l=[&hr4T"Y(AVYGs]mJn#sJl(OFLi0LT-7LeI@L>#[gfD<glB)YS#8Ms<I5b?\UAZmDk,);:HnA$1eV@Z'e?k%X<WOuP-mRAZa6;AN*bA"H.Q=U6<DpZSO&$<_!(0i+t8TDE1K>S14=P`WV"NURGj<QLA#<JK6G22/nI%s36/&W?+i\u8g.Zeh+aQs?BiLW!PCMso!eM794q=b7]uU%.Zj%9a=UW9(HiR(eikc!ola\sHm:CPlkdQb<-##q9k[67:<S$nHBrR9Pbad(r]%>m.]Z&+/+HK&\D;Np-"#0KLuP(*56;n<9G'@h+ZA]c=85'K3EVVX[Nb5bW9o@HcO60DPP$_e6?/FYCm4_AXR9#tl`Og=Amt.!XcM9Hjg22gIY^0b7s9#\8iYrCQ%@>@`q&J]+`;$`@7k^u#GS\_un/WuQ@G@4$E+AgVV_RD^F(r0pd2?d[K.!egTm96?(5KPV1g4iqncm%)fKO(fL;.24%jR/F"TJ,T<4PKu9!_BCBgXWaO.11)/gO[/<@o<*\:E)0UN,Y.8I+BDhC3J0Y_*!H<;(SC:=:@0Q@B2JlfS/fQ:p!2JVAX>`5%4,RteuEU%BrZ-iM[D9*08G4LWE5(@)a#dB-"]>=@*XNJ!/Y=!-9=L4_I'4AU,)r_RJM/_AIlIglm99(R.dE$)o+_uH!@>>]</]FeX0VhnLIot$:Q/i%>R'ihNc2qSkV1Rnd:*f1ANn;*ZB1FjJ*jkXJKS#nk&1O]P8./]$0*q9d.:3F]l)Og9LB^nEOgqO?AQ>"'s'HJP,e#k?\L6J4fZj76Nh]%PPGoHH+,p/o&T_1O)iq=(V?+J9jEa%mVG473+1/;",8r'^JOol<1h'C>erVBD.q6n*73?)mR>e!!]b@34_Z?%>Bu"qE%fo(PrBm]K,2;`2>o\RFD8fX!`5Pjpd14m5/.S*D=DY.FN+uIWh%FO%(UFA6\*o,N[7aPX"b;rle%21G@k9-7D+%_]PWN-]Gg)m1ulo<U^Vo34%j$Pg9?2>I:CL"Fi^t$]@/$45PRZPt/L^`nC8ifsdug6gp!R+dT(0OPl$kmn7hF.GB\JT4Pr.oN?QmkU2,K;tp,AdmTf3k\l&+EEcrG)h%%\>7fs^UJDkS$*8@Z_<kee"tkSLdchmA$%egt?q/`i5lrWA[&CGq%cmqgknKM`<J1cW6&oc8MJd-SM2DWl`OZXJU)DiFU=tEt%F?a.\)%,l4p9)-<4`WFDg58JUE<bCKI0'NYa5\4,qr6i$?\[ER!q<s73WtQ-R0*``mJQbZ=MZ6[!MU&8s^[eBum(M@0_bQrpE#7Xoo@ajE^>jm=R:q.-.*OG3Xg-Z8\!J/O(dD*$&5?J2j4/kBQhO4^QIIq?,5',6RU$EsiJZ'.CpTG0Zg1aQTkZZK6"e&l@==a7[)=+-H_*+_MR"ii2GdI;>\IdQP\"nO<[-VJtZoOCT%\%bupuM:*1\(s8d+agnIF3IAqq@=D.2o1ME'756L<aB@NDq^PmbO,%KRAU8t(a5DsN2V>q01kO<BUSo:Ac+B]T69^Sl84%o\(6u@V-eb;*4WM69rt5/\DJfJFEob47X(*+A9DhF$0[GYV$YmVW<.Y/$g-jR\^,2e>k[$dd\c2rgD;C`Aq@9FR:3uh<iROjVq=/5>G?V*pS5u7gZG8d]RSXam``o2P'@/:FBW]fdYr,2YFj6FX5Yd=QofF:2k;rV?n7k@^+28)%Lg!GF]MdQ03iP7d&.pJX579^`]pXK%fR;^qfC0E<.ZZ4Wm_-G!s,l=6\b)%Yl;OZJ''/*aW#qd-2"<R$BVAX+'[Z>0_"OB>T+B0kD$-l^[DJ\S"YSn?iN^us8`'l9]oL6il*=WP"\spqI,`5G'fGhroRVWSm:;eX4Xm_?bgZ@97qA0YVo_^V1e^'LJ>E!FKL$<FIu%b0433RsDa_l-pEq[UX2Kh.IHL1pb^Y^`,X>C^;:@BLEG/KnM[BTN7.jkC8r<4ebo!55/(AHe\otkiQFu<K09"tYqPl;o$*($;HgmsOaD<DFO<tVs--j<(>OEUtH:;k,3)XR%L5PB$J^Cj"C+i+:6J5T4TdTX-d689f4<[`u^VS[b[4.%qI60/D!p]lKlA#m8-n:m@(uqEka"2`qKgbDP8%R9tH<Br_c`<A8Kbh[C'-s5J.Yo$_h1DNiTI<0X>1nJDgD&r_#)+G?6?"6>PoKcR"ehB1Sp8_7;[N?#?]0PcK*^4@X\[jYKX:7MRs3%^JD&ua[)s!NH)X.rE`51oeC8QT>^b*`_&/&C]iR'*5ZZbD+fHA;!1Lfob@dq\q%?N]M+5b+X+aBsWttpiTut>N,H`*t9qJ=?/A(RY5lmS=%^8rE`=5'?_!N#][],sIf8+"CBT'RXO`eXd@+]S:=ApLPkG6HOD=-+D@gS1G.Nbk]<>[0BCPYm%pk`U2!4]2SkDLuFiqP)15<@JahpfEtiFVPWfnhTDTfm<$=T86O]gt0:No>#GSL\sf-[$$*[l)b2g$FObWD<#/q]F'IePRHWqk/,lBiC6`G-DWN`"3W.'Z%It4+Q_OjT_2LqW'Bb%24o]O9Ho^!ZY5-6#^nMiVT?eNZgknrr^1`a]\~>endstream
+endobj
+13 0 obj
+<<
+/Filter [ /ASCII85Decode /FlateDecode ] /Length 3427
+>>
+stream
+Gat%'gMYe+&q/)-&D.e3&9:3;f6p;ta^>lUY`fo+/<C\$c+YH<aXa@Hj8&Nt>?-o6CgI5h1!Mm^,S*OjU"Gk\qM/GW^HWA%m[!eL,:?`XgF_A59?HQKo&9;ms'B\1VA0c?89.aG!Ssq*\5a^[EdPGiJ%FjCN1/hue*k@P`lMq9qa5iYVN5WO2@Xs3jU;%ZmIi?HDbe3KHIAkQe;OJi%>rYG34-:eO^kjt%7l-2)tNs)DAm]uo3m#6)RWN2VXC(TQLEP=]iS0\J,P*aWrB4cW"$:a)AV(O4Z*[Hn*38%N9!3G]6STtI/=;^q0mD?AH*be^Y<YpV3YcThl<S6[J"^T;uNAe\8Bao8B/=e6[aXP&bT6bVg?j&FY28d[Th:_FsNQ'eE]*M%;Hm>OL?+MTH<B)"*)S-m.G2E`m3V=Dd`@LhLaYXcX4+(01t"(,l)aNEjHd@D:m4`ClI&%pYGA5;0r/DVj@83DU]cROgmqa::Nu:-mfR0#sjq,F&?J$akS_si2r<ddZhhiTIJ2E;.TQ^en':A+Ne<s<i)<&3FPpbPtCqd&6W1>&NtWIcEtaGl:fAHTJM)3OH@c[q9lEgb3'-=l/^0WG-*QG^6!k?=U+RjDs;#-=^[;d>eNX*"SPA5X7o/LJNXXbl9,b_@#m^eHB3^bm#ah>mh)`)HXp[=lHB?W`*[Hh+e:9rLhiP73.uh-(HfX$aqri!5iW7OiVLAqrsXUia0ULnkglI'iogQam6rkYTd!-$-8Q;KDZ7*6pC!'2[')uYM%*g6E-iJF7B._efZuH<56p=2j/L*NpAG]s*&C,2/(8lshUcVFA"a!r8LF"RcMpmemt-iObL=H$5(EI&1Ph+L?b_Pd11O=PbVT(mp,RR>G>2NZU7YL'<SU)/@.mm,L62t$7[[N$+TgBu1<hKo*Om;)KuN;`a\@8Jc]]dMMVF0Tr6.Tnfmtn71kCGo<(I8&iq`u&BC*[J^O>tJ:C$o6aV(#e@/;VgQqcWu\rU"0m$d@DZcZjRkTq%`K&ctc9J-=sOGh<ROdiU4Pm\uj',S'u'pNKW;BNMbH-,Dk>XKHR&bDc5RO>[0oSP]f'<R'$d>GFVXU&/t.uK.IMTeksKaRdVT2pX9SE6$&eq.7Z%Xf"mJCuhd?T1rrg!`\nLBK:=,!ffAk@(f,Q/i_%27$/e!^=J)/Mpm))5%^WU&D:S>-l=Z]6<g5.q"+\nYkcWIJp(<%Qog[!^T]b(FBlIdA-h-``>rZ`Mu$c.RI("3P6Jr.OQn9co7'oR,R5%C".lTl>oT=!l1gp$XAMZYaVmBf8Dqje?j]'oj*qc)%j<-<eM6ILMR^!3oE$*!j,pP_1dqN^l84`d<#8bVumL!MR[nb\-Ph1)9jVg@n[W2g/j"g7B$47&_MW_"9O.8;o%TVFem(Ge;j3qV%*gi'h!/$3UU*/HFZa:6&Y&/j?0ci8osm<Bu8XP7C9`0Pj;#Y!L9WlZE31k0^bc<<g+&;KXD2qLjA3Q_m%a1^uPQ@IGJe>6Yt`uj^S!RCm\Qkj%+V2+(;bejjNsd,JSC;/9;qT<a?&YE5t/s78Wr-fP'%,>3^\AZ%.lZ[KqVk.GI2IMJNYh,#$VZ+Bdt899rb3<\/C'7A)Pkg<49D!8?_l?r)Qd45Xa!W(R?i"ed>nZcEI^JSY#aOVlP91:3uiD3eOZC4?,YS"b9'nI)!;m;[>l3-IN/n`:;^E^:n,gb]kqVVDQ1Z+63,>D2+4C?\Fk/[GH-:=laV^oRO&mlNEmd%L@('A`<R:AY3]"AgW.!ubXde,h=m=qP%l=k+tE*KRe.<a\%2AJIH^a%F<@/V)-;k/hSCl6CFC!4@2C/E=Q%%.Z)?`o^_$lG@-@:r@/.eA+Ba`k<EZl^ik**?-,7Z<Ln!epD4`p_[TT)eKkJZ-:puKrl,a\noW^Te*a/#+7'ie,,L9cq_!KmnNs*p@Y:E;jS'b>Jp]OS-b_s,22.H0FL6QC_gV!(q*[2Kn<leK,G%3/9U`"i&:mjmlNj$:g`sSmWHM71^.8FO,IPa/>.X>!\Tqp4M!GW:gH;=//Mr'1@CMa0Um[+pRNNO88Yhkqh&WRH>3tOWZQ>gi&si>JoMLb>XB.&lUj"72o9f<>GC5>h8#,,cn^^G$40JbGGKa32=?_>B+7"Jh1iL/J6Ws.dst\n7Yk<[Y!01<,OALc"%2MK$pZGjK"4c++I@`<pXNm^)S=I9:EQ2)qPXcZQO&\rj2)^r$YYlL=`ONZ9eML0ck'q%\Xf%U>XK0*3;-ciQWN,<I3fe`RT!GboWD9\:[Jq1"WE9:Oc/E7\VQ.if<-`B\Hp`'+U\tr#VB3AE/hYM$"nLsndIgFLZ-UfjtbKZ*lhKYThm#HJ9GKVQJHRe>XNmlkouAhhN.cMV(L*-#oV(J"duP^6dT-s&)1#M\)!(U"G#3R9ftdbOP5(f4@=LTi<#6b<X)e]6;(@*iu-19kjapO#\m(jVYZUIqfL=6!p!s5YK7=l5&3Ps1pL2*I:f/ZP>71#>X<MbF!\c?/6m"Q^A87(dtu`5U#/Z_*b^gY0?OhB6*"6,`=O,$Y7f/c=V6I!(+Pk:!49Uk"mIP_&6+_HnSg_-JT(nr[h8?s^_[,kRTOJ'arhI@b6,99[P2d;IRpW8)3b%J"ud<nY:r>j[VkOE`mO5NH5NSR#,mNkjT3jXEKqF$U8^k].l08@bk]i@;N&ndA_p?T]].m#bQ'Ghe\aqe]np+QH7\'5a2D[0=,E4.!`p7FLETQBf>&7jC!RnPDqKZb$IW"0!TpOL$pX!Z3K3)M\4^:?1u`8j`+T`aIMM/m<Y8VbiFS"+V1JBPLm6+`,#+$59a0c3^$Mk][XFm>_f4L1(EP2QKnIpoB,4QgAa;Z_(0ti%1@^l_-VL9/%B^X-E:s`Lb5X*gJiS%9WhnII3u\82]Z#jU<5qp@Ea='UnZT_XI$!0&(m?I\'$+&2IY-m+!Z\jX0_X<^?LMtBiR(:DW?SLT&hD9:UJ%Z`MFK"b![QChp-sEXBmb%<"I[a;cR3t6k^nEtNHLpl>cnH+d?R[^\PqouMfkXM.pX'kq#\mWoY2KO=I/G+kN&h]^]+(eV;uI-Zu^o4jegjqDjUIAp*NjhaT@0g^Eda3lXTg7g\,\CpZja2S+#lT^M2I%En50_[QgDID;)E_S]<U)gFq`0:##C5`OL:MDR(*<RE%0lTYZ*+[f-"aY%cYiC`_38D82U]2/V)$k&F,Mp*Yc<PBgF)V&&DVD;/FAj-mFDcfT4Ag\.Q:\>F-?`B:;;g\t_<DDI8g^1`,sVJZ3gH!Jh,rV8RYC"rW@?1N\TnV(9=KjF5&?1EVS/^.s<<=>R*f[Rd6Q56k^W5ZJ'[WSjt`?K:Y_6p$nj'<Sp1W"DD:X@NoI$0De?iOI)<p!VSQi9Wnr^<P2BUW+_:J-"$Vq0k\^%fn]%9**+TDX06+g:kU$+Q6H~>endstream
+endobj
+xref
+0 14
+0000000000 65535 f 
+0000000061 00000 n 
+0000000112 00000 n 
+0000000219 00000 n 
+0000000331 00000 n 
+0000000450 00000 n 
+0000000645 00000 n 
+0000000840 00000 n 
+0000001035 00000 n 
+0000001104 00000 n 
+0000001384 00000 n 
+0000001456 00000 n 
+0000004182 00000 n 
+0000007449 00000 n 
+trailer
+<<
+/ID 
+[<112ce70afde50720dbcbf78f979ec59c><112ce70afde50720dbcbf78f979ec59c>]
+% ReportLab generated PDF document -- digest (opensource)
+
+/Info 9 0 R
+/Root 8 0 R
+/Size 14
+>>
+startxref
+10968
+%%EOF

analysis/out/lcz_class_stats.txt

@@ -0,0 +1,30 @@
+LCZ Class Distribution Statistics
+============================================================
+
+Image dimensions: 45,096 x 18,379
+Total pixels (excluding class 0): 3,978,683
+
+Dominant class: 6 (Open low-rise)
+
+Class Ranking by Area Coverage:
+------------------------------------------------------------
+Rank  Class       Count    Percentage  Description
+------------------------------------------------------------
+1     6             2,304,386    57.9183%  Open low-rise
+2     14              435,041    10.9343%  Low plants
+3     11              286,834     7.2093%  Dense trees
+4     16              198,951     5.0004%  Bare soil
+5     13              183,003     4.5996%  Bush, scrub
+6     8               135,060     3.3946%  Large low-rise
+7     17              128,621     3.2328%  Water
+8     12              128,592     3.2320%  Scattered trees
+9     15               77,947     1.9591%  Bare rock
+10    18               56,554     1.4214%  Custom/Unknown
+11    2                17,578     0.4418%  Compact mid-rise
+12    3                10,448     0.2626%  Compact low-rise
+13    10                6,202     0.1559%  Heavy industry
+14    4                 4,514     0.1135%  Open high-rise
+15    5                 2,839     0.0714%  Open mid-rise
+16    1                 2,113     0.0531%  Compact high-rise
+
+============================================================

analysis/out/lst_per_lcz_stats.txt

@@ -0,0 +1,26 @@
+LST Distribution per LCZ Class
+======================================================================
+
+Total LST files processed: 1,811,272
+LCZ classes with data: 16
+
+Class   Description                       Pixels      Mean       Std    Median     Min     Max
+----------------------------------------------------------------------------------------------
+LCZ 1   Compact high-rise              8,639,972     75.89     26.96        79    -189     148
+LCZ 2   Compact mid-rise              77,403,212     77.34     26.79        81    -189     175
+LCZ 3   Compact low-rise              47,478,288     78.88     29.14        83    -189     151
+LCZ 4   Open high-rise                20,344,389     82.40     28.95        85    -189     164
+LCZ 5   Open mid-rise                 16,910,931     89.08     28.14        91    -189     161
+LCZ 6   Open low-rise             11,185,948,355     82.78     28.56        85    -189     192
+LCZ 8   Large low-rise               737,551,075     94.14     27.95        98    -189     187
+LCZ 10  Heavy industry                28,123,635     82.89     31.39        85    -189     164
+LCZ 11  Dense trees                1,234,477,681     67.21     25.56        71    -189     211
+LCZ 12  Scattered trees              583,568,105     73.39     27.82        77    -189     211
+LCZ 13  Bush, scrub                1,302,142,502    100.87     27.01       103    -189     179
+LCZ 14  Low plants                 2,372,490,276     82.20     29.32        84    -189     211
+LCZ 15  Bare rock                    501,687,608     97.59     26.48       100    -189     177
+LCZ 16  Bare soil                  1,654,509,721    103.50     27.53       108    -189     181
+LCZ 17  Water                        514,340,946     65.40     25.38        68    -189     169
+LCZ 18  Custom/Unknown               194,038,393     97.10     26.06        99    -189     153
+
+Total                             20,479,655,089

analysis/out/lst_temperature_stats.txt

@@ -0,0 +1,424 @@
+LST Temperature Distribution Statistics
+==================================================
+
+Total files processed: 1,811,272
+Total pixels (excluding 0°F): 22,134,369,574
+
+Min temperature: -189°F
+Max temperature: 211°F
+Mean temperature: 84.27°F
+Median temperature: 85°F
+Mode temperature: 89°F
+Std deviation: 29.61°F
+
+Percentiles:
+  10th: 51°F
+  25th: 66°F
+  50th: 85°F
+  75th: 104°F
+  90th: 121°F
+
+==================================================
+Full Histogram Data
+Temp(°F)	Count		Probability
+--------------------------------------------------
+-189		36,251,640		0.00163780
+-188		35,439		0.00000160
+-187		35,904		0.00000162
+-186		37,488		0.00000169
+-185		38,206		0.00000173
+-184		39,368		0.00000178
+-183		40,322		0.00000182
+-182		40,801		0.00000184
+-181		41,656		0.00000188
+-180		42,985		0.00000194
+-179		43,742		0.00000198
+-178		45,633		0.00000206
+-177		46,522		0.00000210
+-176		47,888		0.00000216
+-175		49,239		0.00000222
+-174		50,420		0.00000228
+-173		51,431		0.00000232
+-172		52,392		0.00000237
+-171		53,077		0.00000240
+-170		53,560		0.00000242
+-169		55,053		0.00000249
+-168		56,262		0.00000254
+-167		56,998		0.00000258
+-166		57,735		0.00000261
+-165		58,097		0.00000262
+-164		58,656		0.00000265
+-163		60,379		0.00000273
+-162		61,254		0.00000277
+-161		62,235		0.00000281
+-160		63,203		0.00000286
+-159		62,793		0.00000284
+-158		64,597		0.00000292
+-157		65,391		0.00000295
+-156		65,579		0.00000296
+-155		65,337		0.00000295
+-154		65,987		0.00000298
+-153		66,491		0.00000300
+-152		67,515		0.00000305
+-151		66,940		0.00000302
+-150		67,395		0.00000304
+-149		68,243		0.00000308
+-148		69,925		0.00000316
+-147		70,823		0.00000320
+-146		72,184		0.00000326
+-145		72,311		0.00000327
+-144		73,688		0.00000333
+-143		76,386		0.00000345
+-142		78,552		0.00000355
+-141		80,552		0.00000364
+-140		82,697		0.00000374
+-139		84,048		0.00000380
+-138		84,453		0.00000382
+-137		87,801		0.00000397
+-136		89,880		0.00000406
+-135		93,504		0.00000422
+-134		96,656		0.00000437
+-133		100,018		0.00000452
+-132		100,857		0.00000456
+-131		102,183		0.00000462
+-130		104,170		0.00000471
+-129		105,650		0.00000477
+-128		107,247		0.00000485
+-127		106,616		0.00000482
+-126		109,141		0.00000493
+-125		110,852		0.00000501
+-124		112,902		0.00000510
+-123		112,019		0.00000506
+-122		114,243		0.00000516
+-121		113,886		0.00000515
+-120		116,808		0.00000528
+-119		119,117		0.00000538
+-118		120,144		0.00000543
+-117		122,225		0.00000552
+-116		125,102		0.00000565
+-115		127,927		0.00000578
+-114		130,343		0.00000589
+-113		133,200		0.00000602
+-112		135,114		0.00000610
+-111		137,823		0.00000623
+-110		136,479		0.00000617
+-109		139,720		0.00000631
+-108		137,158		0.00000620
+-107		139,854		0.00000632
+-106		144,329		0.00000652
+-105		146,899		0.00000664
+-104		150,054		0.00000678
+-103		151,248		0.00000683
+-102		155,009		0.00000700
+-101		157,033		0.00000709
+-100		161,707		0.00000731
+-99		162,157		0.00000733
+-98		166,849		0.00000754
+-97		167,537		0.00000757
+-96		171,898		0.00000777
+-95		175,085		0.00000791
+-94		179,598		0.00000811
+-93		180,098		0.00000814
+-92		182,006		0.00000822
+-91		187,332		0.00000846
+-90		190,658		0.00000861
+-89		197,191		0.00000891
+-88		196,729		0.00000889
+-87		202,424		0.00000915
+-86		202,238		0.00000914
+-85		209,128		0.00000945
+-84		213,288		0.00000964
+-83		222,725		0.00001006
+-82		227,868		0.00001029
+-81		237,465		0.00001073
+-80		251,257		0.00001135
+-79		260,433		0.00001177
+-78		272,390		0.00001231
+-77		290,171		0.00001311
+-76		303,317		0.00001370
+-75		313,309		0.00001415
+-74		322,618		0.00001458
+-73		334,632		0.00001512
+-72		353,130		0.00001595
+-71		359,524		0.00001624
+-70		357,941		0.00001617
+-69		351,211		0.00001587
+-68		350,624		0.00001584
+-67		349,559		0.00001579
+-66		334,525		0.00001511
+-65		326,821		0.00001477
+-64		322,808		0.00001458
+-63		323,929		0.00001463
+-62		325,730		0.00001472
+-61		323,769		0.00001463
+-60		322,102		0.00001455
+-59		328,793		0.00001485
+-58		332,627		0.00001503
+-57		340,329		0.00001538
+-56		340,596		0.00001539
+-55		341,012		0.00001541
+-54		347,042		0.00001568
+-53		355,861		0.00001608
+-52		360,641		0.00001629
+-51		358,276		0.00001619
+-50		362,363		0.00001637
+-49		367,568		0.00001661
+-48		390,242		0.00001763
+-47		404,833		0.00001829
+-46		424,023		0.00001916
+-45		436,665		0.00001973
+-44		460,423		0.00002080
+-43		478,526		0.00002162
+-42		504,467		0.00002279
+-41		530,549		0.00002397
+-40		552,373		0.00002496
+-39		566,420		0.00002559
+-38		578,391		0.00002613
+-37		588,748		0.00002660
+-36		594,086		0.00002684
+-35		625,401		0.00002825
+-34		652,510		0.00002948
+-33		657,182		0.00002969
+-32		630,716		0.00002849
+-31		635,643		0.00002872
+-30		639,302		0.00002888
+-29		657,737		0.00002972
+-28		672,061		0.00003036
+-27		690,315		0.00003119
+-26		701,577		0.00003170
+-25		712,666		0.00003220
+-24		775,338		0.00003503
+-23		894,227		0.00004040
+-22		1,027,456		0.00004642
+-21		1,107,940		0.00005006
+-20		1,144,083		0.00005169
+-19		1,168,537		0.00005279
+-18		1,144,842		0.00005172
+-17		1,154,787		0.00005217
+-16		1,210,320		0.00005468
+-15		1,206,142		0.00005449
+-14		1,197,578		0.00005410
+-13		1,207,750		0.00005456
+-12		1,219,800		0.00005511
+-11		1,412,073		0.00006380
+-10		1,842,731		0.00008325
+-9		1,935,957		0.00008746
+-8		1,760,387		0.00007953
+-7		1,783,568		0.00008058
+-6		1,738,855		0.00007856
+-5		1,666,534		0.00007529
+-4		1,604,878		0.00007251
+-3		1,683,082		0.00007604
+-2		1,908,774		0.00008624
+-1		2,295,132		0.00010369
+1		5,288,349		0.00023892
+2		5,363,497		0.00024232
+3		5,104,251		0.00023060
+4		5,120,197		0.00023132
+5		6,102,346		0.00027570
+6		6,687,498		0.00030213
+7		6,435,962		0.00029077
+8		6,472,710		0.00029243
+9		6,525,365		0.00029481
+10		6,434,344		0.00029069
+11		6,754,415		0.00030516
+12		7,826,022		0.00035357
+13		8,929,130		0.00040341
+14		9,645,440		0.00043577
+15		9,948,900		0.00044948
+16		10,493,584		0.00047409
+17		10,831,234		0.00048934
+18		11,395,308		0.00051482
+19		12,091,541		0.00054628
+20		14,048,469		0.00063469
+21		16,700,032		0.00075448
+22		18,984,842		0.00085771
+23		20,147,106		0.00091022
+24		20,985,736		0.00094811
+25		21,421,466		0.00096779
+26		22,748,678		0.00102775
+27		24,605,197		0.00111163
+28		26,070,994		0.00117785
+29		28,734,290		0.00129818
+30		31,255,320		0.00141207
+31		36,587,154		0.00165296
+32		43,714,099		0.00197494
+33		47,944,312		0.00216606
+34		51,763,802		0.00233862
+35		55,332,810		0.00249986
+36		57,005,442		0.00257543
+37		59,242,241		0.00267648
+38		62,218,734		0.00281096
+39		67,461,422		0.00304781
+40		72,502,195		0.00327555
+41		77,665,614		0.00350882
+42		85,132,479		0.00384617
+43		90,827,939		0.00410348
+44		98,760,930		0.00446188
+45		106,472,252		0.00481027
+46		114,927,802		0.00519228
+47		121,931,391		0.00550869
+48		132,489,709		0.00598570
+49		142,506,828		0.00643826
+50		154,691,491		0.00698875
+51		163,021,106		0.00736507
+52		172,606,469		0.00779812
+53		180,839,336		0.00817007
+54		187,992,665		0.00849325
+55		196,754,265		0.00888908
+56		199,502,795		0.00901326
+57		210,031,648		0.00948894
+58		218,436,703		0.00986867
+59		228,996,685		0.01034575
+60		230,665,178		0.01042113
+61		237,091,158		0.01071145
+62		244,724,710		0.01105632
+63		255,201,905		0.01152967
+64		261,793,910		0.01182748
+65		271,549,964		0.01226825
+66		278,993,685		0.01260455
+67		281,980,067		0.01273947
+68		285,372,456		0.01289273
+69		283,307,887		0.01279946
+70		285,363,255		0.01289231
+71		284,645,948		0.01285991
+72		286,792,859		0.01295690
+73		285,792,272		0.01291170
+74		289,034,733		0.01305819
+75		286,985,938		0.01296563
+76		288,909,317		0.01305252
+77		289,657,794		0.01308634
+78		293,539,672		0.01326171
+79		293,076,033		0.01324077
+80		296,869,227		0.01341214
+81		300,753,645		0.01358763
+82		302,268,952		0.01365609
+83		304,020,469		0.01373522
+84		301,344,003		0.01361430
+85		303,175,634		0.01369705
+86		304,005,869		0.01373456
+87		308,227,622		0.01392529
+88		308,207,005		0.01392436
+89		311,048,022		0.01405272
+90		308,870,895		0.01395436
+91		309,264,212		0.01397213
+92		304,613,407		0.01376201
+93		303,834,943		0.01372684
+94		301,155,871		0.01360580
+95		295,871,230		0.01336705
+96		293,654,991		0.01326692
+97		286,626,709		0.01294940
+98		282,394,059		0.01275817
+99		274,784,524		0.01241438
+100		270,250,104		0.01220952
+101		261,583,266		0.01181797
+102		256,233,553		0.01157628
+103		248,190,112		0.01121288
+104		243,688,065		0.01100949
+105		236,204,740		0.01067140
+106		231,230,712		0.01044668
+107		224,615,989		0.01014784
+108		217,485,314		0.00982568
+109		213,539,165		0.00964740
+110		207,600,985		0.00937912
+111		204,845,866		0.00925465
+112		199,374,355		0.00900746
+113		196,267,975		0.00886711
+114		190,677,974		0.00861457
+115		186,896,222		0.00844371
+116		180,366,460		0.00814871
+117		175,628,335		0.00793464
+118		168,576,388		0.00761605
+119		163,891,505		0.00740439
+120		158,339,755		0.00715357
+121		152,004,656		0.00686736
+122		148,242,654		0.00669740
+123		142,985,849		0.00645990
+124		138,781,969		0.00626998
+125		132,309,630		0.00597756
+126		127,439,801		0.00575755
+127		121,528,626		0.00549049
+128		117,109,018		0.00529082
+129		110,912,196		0.00501086
+130		105,690,880		0.00477497
+131		99,314,409		0.00448689
+132		94,578,692		0.00427293
+133		88,992,329		0.00402055
+134		83,942,044		0.00379238
+135		78,241,347		0.00353484
+136		71,796,863		0.00324368
+137		65,802,215		0.00297285
+138		58,835,132		0.00265809
+139		52,806,867		0.00238574
+140		45,785,786		0.00206854
+141		39,333,531		0.00177703
+142		32,839,643		0.00148365
+143		27,371,476		0.00123661
+144		22,342,985		0.00100942
+145		18,609,252		0.00084074
+146		15,669,975		0.00070795
+147		13,125,320		0.00059298
+148		10,622,728		0.00047992
+149		8,399,505		0.00037948
+150		6,439,153		0.00029091
+151		4,624,799		0.00020894
+152		3,277,751		0.00014808
+153		2,205,303		0.00009963
+154		1,514,014		0.00006840
+155		994,773		0.00004494
+156		661,861		0.00002990
+157		464,605		0.00002099
+158		344,147		0.00001555
+159		245,319		0.00001108
+160		182,734		0.00000826
+161		136,406		0.00000616
+162		102,601		0.00000464
+163		75,784		0.00000342
+164		54,608		0.00000247
+165		39,544		0.00000179
+166		26,263		0.00000119
+167		19,053		0.00000086
+168		11,656		0.00000053
+169		6,728		0.00000030
+170		3,819		0.00000017
+171		2,571		0.00000012
+172		1,943		0.00000009
+173		539		0.00000002
+174		205		0.00000001
+175		130		0.00000001
+176		102		0.00000000
+177		62		0.00000000
+178		75		0.00000000
+179		59		0.00000000
+180		46		0.00000000
+181		53		0.00000000
+182		49		0.00000000
+183		47		0.00000000
+184		39		0.00000000
+185		31		0.00000000
+186		42		0.00000000
+187		39		0.00000000
+188		24		0.00000000
+189		22		0.00000000
+190		15		0.00000000
+191		27		0.00000000
+192		12		0.00000000
+193		13		0.00000000
+194		17		0.00000000
+195		21		0.00000000
+196		31		0.00000000
+197		15		0.00000000
+198		11		0.00000000
+199		17		0.00000000
+200		10		0.00000000
+201		6		0.00000000
+202		15		0.00000000
+203		14		0.00000000
+204		12		0.00000000
+205		10		0.00000000
+206		11		0.00000000
+207		8		0.00000000
+208		10		0.00000000
+209		11		0.00000000
+210		3		0.00000000
+211		298		0.00000001