Examples¶

import pandas as pd
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
from datetime import datetime

# Load the metadata CSV
df = pd.read_csv('./temperature_analysis.csv')

# Convert date and time columns to datetime
df['datetime'] = pd.to_datetime(df['date'] + ' ' + df['time'])

# Convert camera_temperature to float
df['camera_temperature'] = pd.to_numeric(df['camera_temperature'], errors='coerce')

# Drop rows with missing temperature data
df = df.dropna(subset=['camera_temperature'])

# Sort by datetime
df = df.sort_values('datetime')

# Create a plot
plt.figure(figsize=(12, 6))
plt.plot(df['datetime'], df['camera_temperature'], 'o-', markersize=2)
plt.title('Camera Temperature Over Time')
plt.xlabel('Date and Time')
plt.ylabel('Temperature (°C)')

# Format x-axis as dates
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m-%d'))
plt.gcf().autofmt_xdate()  # Rotate date labels

# Add grid for better readability
plt.grid(True, alpha=0.3)

# Save the plot
plt.tight_layout()
plt.savefig('camera_temperature_trend.png')
plt.close()

print(f"Analysis complete. Results saved to camera_temperature_trend.png")

import argparse
import glob
import os
from datetime import datetime

import matplotlib.dates as mdates
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
from PIL import Image


def analyze_phenocam_data(data_dir, metadata_file):
    # Find all full resolution images
    image_dir = os.path.join(data_dir, "full_res")
    image_files = sorted(glob.glob(os.path.join(image_dir, "*.jpg")))
    print(f"Found {len(image_files)} images to analyze")

    # Load metadata
    metadata_df = pd.read_csv(metadata_file)
    print(f"Loaded metadata with {len(metadata_df)} entries")

    # Create a dictionary for fast lookup of metadata by filename
    metadata_dict = {os.path.basename(row['filename']): row for _, row in metadata_df.iterrows()}

    # Extract greenness index from images and merge with metadata
    data_points = []

    for img_file in image_files:
        filename = os.path.basename(img_file)
        meta_filename = filename.replace('.jpg', '.meta')

        # Skip if we don't have matching metadata
        if meta_filename not in metadata_dict:
            continue

        # Get metadata for this image
        meta = metadata_dict[meta_filename]

        # Parse timestamp
        timestamp = datetime.strptime(f"{meta['date']} {meta['time']}", "%Y-%m-%d %H:%M:%S")

        # Open image and calculate greenness index
        img = Image.open(img_file)
        img_array = np.array(img)
        # Simple greenness index (G - R)
        r_channel = img_array[:, :, 0].mean()
        g_channel = img_array[:, :, 1].mean()
        greenness = g_channel - r_channel

        # Extract camera temperature if available
        camera_temp = meta.get('camera_temperature', None)

        # Store the data point
        data_points.append({
            'timestamp': timestamp,
            'greenness': greenness,
            'camera_temp': camera_temp,
            'filename': filename,
            'day_of_year': meta['doy'],
            'day_of_week': meta['day']
        })

    # Convert to DataFrame
    analysis_df = pd.DataFrame(data_points)

    # Sort by timestamp
    analysis_df = analysis_df.sort_values('timestamp')

    # Plot results using datetime on x-axis
    plt.figure(figsize=(12, 8))

    # Create two subplots
    ax1 = plt.subplot(2, 1, 1)
    ax1.plot(analysis_df['timestamp'], analysis_df['greenness'], 'o-', color='green')
    ax1.set_title("Greenness Index Over Time")
    ax1.set_ylabel("Greenness Index (G-R)")
    ax1.grid(True, alpha=0.3)

    # Add temperature subplot if available
    if 'camera_temp' in analysis_df and not analysis_df['camera_temp'].isna().all():
        ax2 = plt.subplot(2, 1, 2, sharex=ax1)
        ax2.plot(analysis_df['timestamp'], analysis_df['camera_temp'], 'o-', color='red')
        ax2.set_title("Camera Temperature")
        ax2.set_ylabel("Temperature (°C)")
        ax2.set_xlabel("Date Time (UTC)")
        ax2.grid(True, alpha=0.3)

    # Format x-axis to show dates nicely
    plt.gcf().autofmt_xdate()  # Rotate date labels

    # Save plot
    plt.tight_layout()
    output_file = os.path.join(data_dir, "combined_analysis.png")
    plt.savefig(output_file)
    plt.close()

    print(f"Analysis complete. Results saved to {output_file}")


if __name__ == "__main__":
    parser = argparse.ArgumentParser(description="Analyze PhenoCam images with metadata")
    parser.add_argument("--data-dir", required=True, help="Directory containing PhenoCam data")
    parser.add_argument("--metadata", required=True, help="Path to metadata CSV file")
    args = parser.parse_args()
    analyze_phenocam_data(args.data_dir, args.metadata)