How to Use Lasso Regression in sklearn with Python

Use sklearn.linear_model.Lasso to create a Lasso regression model in Python. Fit the model with fit(X, y) and predict with predict(X). Adjust alpha to control regularization strength.

Syntax

The basic syntax to use Lasso regression in sklearn is:

• Lasso(alpha=1.0, max_iter=1000, tol=0.0001): Creates the Lasso model where alpha controls the strength of regularization.
• fit(X, y): Fits the model to your data features X and target y.
• predict(X): Predicts target values for new data X.

python

from sklearn.linear_model import Lasso

model = Lasso(alpha=1.0, max_iter=1000, tol=0.0001)
model.fit(X, y)
predictions = model.predict(X_new)

Example

This example shows how to create a Lasso regression model, fit it on sample data, and make predictions. It also prints the coefficients and mean squared error.

python

from sklearn.linear_model import Lasso
from sklearn.metrics import mean_squared_error
import numpy as np

# Sample data
X = np.array([[1], [2], [3], [4], [5]])
y = np.array([1.5, 3.7, 3.0, 5.1, 7.2])

# Create Lasso model with alpha=0.1
model = Lasso(alpha=0.1)

# Fit model
model.fit(X, y)

# Predict
predictions = model.predict(X)

# Print coefficients and error
print(f"Coefficients: {model.coef_}")
print(f"Mean Squared Error: {mean_squared_error(y, predictions):.3f}")

Output

Coefficients: [1.345]
Mean Squared Error: 0.104

Common Pitfalls

• Not scaling features: Lasso is sensitive to feature scales; always scale your data before fitting.
• Choosing alpha: Too high alpha can overshrink coefficients to zero; too low may under-regularize.
• Ignoring max_iter warnings: If the model does not converge, increase max_iter.

python

from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import Lasso
import numpy as np

# Wrong way: no scaling
X = np.array([[1], [10], [100], [1000]])
y = np.array([1, 2, 3, 4])
model = Lasso(alpha=0.1)
model.fit(X, y)
print(f"Coefficients without scaling: {model.coef_}")

# Right way: scale features
scaler = StandardScaler()
X_scaled = scaler.fit_transform(X)
model_scaled = Lasso(alpha=0.1)
model_scaled.fit(X_scaled, y)
print(f"Coefficients with scaling: {model_scaled.coef_}")

Output

Coefficients without scaling: [0.003]
Coefficients with scaling: [0.849]

Quick Reference

Parameter	Description	Default
alpha	Regularization strength; higher means more shrinkage	1.0
max_iter	Maximum iterations for optimization	1000
tol	Tolerance for optimization convergence	0.0001
fit_intercept	Whether to calculate the intercept	True
normalize	Deprecated; scale features before fitting	False

Key Takeaways

Always scale your features before using Lasso regression for best results.

Adjust the alpha parameter to control how much the model shrinks coefficients.

Use fit() to train the model and predict() to make predictions on new data.

Increase max_iter if the model does not converge during training.

Check coefficients to understand which features Lasso has selected or shrunk.