2

I am trying to run Linear Regression with LASSO using Python's Scikit-learn package.

For Lasso, my configurations are as follows:

lasso_eps = 0.0001
lasso_alpha = 20
lasso_iter = 5000

And the code for the model is as follows:

lasso_cv = LassoCV(eps=lasso_eps, n_alphas=lasso_alpha, max_iter=lasso_iter, normalize=True, cv=5)
model = make_pipeline(PolynomialFeatures(degree=2, interaction_only=False), lasso_cv)
model.fit(X_train, y_train.values.ravel())

y_predict = model.predict(X_test)
model_score = model.score(X_test, y_test)
print(f"Accuracy: {model_score}")

Though, I am getting my model score properly; I am getting the following warning as well:

/Users/pankajkumar/anaconda3/lib/python3.6/site-packages/sklearn/linear_model/coordinate_descent.py:491: ConvergenceWarning: Objective did not converge. You might want to increase the number of iterations. Fitting data with very small alpha may cause precision problems.
  ConvergenceWarning)

Could anyone suggest, what this warning means and how it can be resolved? Any help would be much appreciated.

PankajK
  • 187
  • 3
  • 14
  • Adding the following `ignore_warnings` helped in removing the warning: `with ignore_warnings(category=ConvergenceWarning):` But still, I am curious, what this meant. – PankajK Mar 19 '18 at 01:26

0 Answers0