
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.model_selection import train_test_split
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import mean_absolute_error, mean_squared_error, r2_score
from sklearn.impute import SimpleImputer

# Load the dataset
file_path = 'Gold Price Prediction.csv'
data = pd.read_csv(file_path)

# Data Preprocessing

# Convert 'Date' to datetime format
data['Date'] = pd.to_datetime(data['Date'])

# Drop rows with missing target values (Price Tomorrow)
data_clean = data.dropna(subset=['Price Tomorrow'])

# Handle remaining missing values by forward-filling
data_clean.fillna(method='ffill', inplace=True)

/tmp/ipykernel_104211/1417853290.py:16: UserWarning: Could not infer format, so each element will be parsed individually, falling back to `dateutil`. To ensure parsing is consistent and as-expected, please specify a format.
  data['Date'] = pd.to_datetime(data['Date'])
/tmp/ipykernel_104211/1417853290.py:22: SettingWithCopyWarning: 
A value is trying to be set on a copy of a slice from a DataFrame

See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy
  data_clean.fillna(method='ffill', inplace=True)


# Define the features (X) and the target (y)
X = data_clean.drop(columns=['Date', 'Price Tomorrow'])
y = data_clean['Price Tomorrow']

# Impute missing values with the mean for numerical columns
imputer = SimpleImputer(strategy='mean')
X_imputed = imputer.fit_transform(X)

# Split the imputed data into training and testing sets
X_train, X_test, y_train, y_test = train_test_split(X_imputed, y, test_size=0.2, random_state=42)


# Plot the gold price over time
plt.figure(figsize=(10, 6))
plt.plot(data_clean['Date'], data_clean['Price Today'], label='Gold Price Today')
plt.xlabel('Date')
plt.ylabel('Price')
plt.title('Gold Price Over Time')
plt.legend()
plt.grid(True)
plt.show()


# Correlation heatmap
plt.figure(figsize=(12, 8))
sns.heatmap(data_clean.corr(), annot=True, cmap='coolwarm', fmt='.2f')
plt.title('Correlation Matrix')
plt.show()


# Train a Random Forest model for prediction
model = RandomForestRegressor(n_estimators=100, random_state=42)
model.fit(X_train, y_train)

# Predict on test set and evaluate
y_pred = model.predict(X_test)
mae = mean_absolute_error(y_test, y_pred)
mse = mean_squared_error(y_test, y_pred)
r2 = r2_score(y_test, y_pred)

# Display the Mean Absolute Error
print(f"Mean Absolute Error: {mae}")
print(f"Mean Squared Error: {mse}")
print(f"R^2 Score: {r2}")

Mean Absolute Error: 8.487179259259312
Mean Squared Error: 172.1552084773367
R^2 Score: 0.9959469799841132


import matplotlib.pyplot as plt
import numpy as np

# 1. Visualizing Model Accuracy

# Plot Predicted vs Actual values
plt.figure(figsize=(10, 6))
plt.scatter(y_test, y_pred, alpha=0.7, color='b')  # Scatter plot
plt.plot([min(y_test), max(y_test)], [min(y_test), max(y_test)], color='r', lw=2)  # Line for perfect prediction
plt.xlabel('Actual Prices')
plt.ylabel('Predicted Prices')
plt.title('Predicted vs Actual Gold Prices')
plt.grid(True)
plt.show()

# Plot residuals (errors)
residuals = y_test - y_pred
plt.figure(figsize=(10, 6))
plt.scatter(y_pred, residuals, alpha=0.7, color='g')
plt.axhline(y=0, color='r', linestyle='--')
plt.xlabel('Predicted Prices')
plt.ylabel('Residuals (Errors)')
plt.title('Residuals Plot')
plt.grid(True)
plt.show()

# 2. Feature Importance

# Get feature importances from the Random Forest model
importances = model.feature_importances_
feature_names = X.columns

# Sort the feature importance values
indices = np.argsort(importances)

# Plot feature importances
plt.figure(figsize=(10, 6))
plt.title('Feature Importance')
plt.barh(range(len(indices)), importances[indices], color='b', align='center')
plt.yticks(range(len(indices)), [feature_names[i] for i in indices])
plt.xlabel('Relative Importance')
plt.show()

Gold Price Prediction¶

Data Preprocessing¶

Train Test Split¶

Prediction¶

Conclusion¶