This is a starter notebook for the Conch Republic Coral Quest. It is designed to help you kickstart your quest.

coral-quest-starter.txt

# Coral Quest - Starter Notebook

"""
🧭 ABOUT THIS NOTEBOOK

This is a **starter notebook** for the Conch Republic Coral Quest. It is designed to help you get up and running with:
- Loading and exploring the CREMP coral reef dataset
- Visualizing trends over time and by site
- Building simple forecasting and regression models

Feel free to **modify, extend, and improve** this notebook based on your own ideas and analytical approach.
Some things you might want to change or add:
- Use more features (e.g. bleaching, depth, pollution, etc.)
- Expand to multiple reef sites
- Try clustering, time-lagged regression, or other ML models
- Refine evaluation metrics

Good luck, and don’t forget to tell a story with your data! 🌊📊
"""

# 📦 1. Import Libraries
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import numpy as np
from prophet import Prophet
from sklearn.ensemble import RandomForestRegressor
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error

sns.set(style="whitegrid")

# 🧪 2. Load Dataset
# TODO: Replace this with actual dataset file path or URL
data = pd.read_csv("coral_data.csv")  # placeholder filename

# 🔍 3. Preview Data
print("\nFirst few rows of the dataset:")
print(data.head())
print("\nColumns:", data.columns)

# 🧼 4. Basic Cleanup
# Optional: Handle missing values, data types, column renaming if needed
# For example:
# data = data.dropna()

# 📈 5. Trend Over Time (Coral Cover)
plt.figure(figsize=(12, 6))
sns.lineplot(data=data, x='Year', y='Stony_Coral_Cover', hue='Site')
plt.title('Stony Coral Cover Over Time by Site')
plt.ylabel('Stony Coral Cover (%)')
plt.xlabel('Year')
plt.legend(title='Reef Site')
plt.tight_layout()
plt.show()

# 🗺️ 6. Correlation Heatmap
plt.figure(figsize=(6, 4))
corr = data[["Stony_Coral_Cover", "Species_Richness", "Temperature"]].corr()
sns.heatmap(corr, annot=True, cmap='coolwarm')
plt.title('Correlation Matrix')
plt.tight_layout()
plt.show()

# 📊 7. Forecasting with Prophet (Single Site Example)
# Prepare data for Prophet (must have 'ds' and 'y' columns)
site_data = data[data['Site'] == data['Site'].unique()[0]]  # first site
prophet_df = site_data[['Year', 'Stony_Coral_Cover']].rename(columns={'Year': 'ds', 'Stony_Coral_Cover': 'y'})
prophet_df['ds'] = pd.to_datetime(prophet_df['ds'], format='%Y')

model = Prophet()
model.fit(prophet_df)

# Create future dataframe
future = model.make_future_dataframe(periods=5, freq='Y')
forecast = model.predict(future)

# Plot forecast
model.plot(forecast)
plt.title('Forecast of Coral Cover for Site A (Prophet)')
plt.show()

# 🧠 8. Regression Model: Random Forest
# Select features and target
features = data[["Year", "Temperature"]]
target = data["Stony_Coral_Cover"]

# Split data into train and test sets
X_train, X_test, y_train, y_test = train_test_split(features, target, test_size=0.2, random_state=42)

# Initialize and train the model
rf = RandomForestRegressor(n_estimators=100, random_state=42)
rf.fit(X_train, y_train)

# Predict and evaluate
y_pred = rf.predict(X_test)
mse = mean_squared_error(y_test, y_pred)
print(f"\nRandom Forest MSE: {mse:.2f}")

# Plot predictions vs true values
plt.figure(figsize=(6, 4))
plt.scatter(y_test, y_pred)
plt.plot([y_test.min(), y_test.max()], [y_test.min(), y_test.max()], 'r--')
plt.xlabel("Actual Coral Cover")
plt.ylabel("Predicted Coral Cover")
plt.title("Random Forest Prediction Accuracy")
plt.tight_layout()
plt.show()

# ✅ Notebook Complete! Add more sites, models, or pollution variables as needed.