Example

Simple example for parametric_si

# import modules
import parametric_si as psi
import numpy as np

# generate data
X = np.random.randn(100,10)
beta = np.array([0,0,0,0,0,0,0,0,0,0])
y = X @ beta + np.random.randn(100)

# execute selective inference
result_lasso  = psi.parametric_lasso_si(X,y,α)
result_stepwise = psi.parametric_sfs_si(X,y,k)
result_lars = psi.parametric_lars_si(X,y,k)

# execute selective inference with cv
result_lasso_cv = psi.parametric_lasso_cv_si(X,y,[0.01,0.1,1,10,100],5)
result_stepwise_cv = psi.parametric_sfs_cv_si(X,y,[1,2,3,4,5],5)
result_lars_cv = psi.parametric_lars_cv_si(X,y,[1,2,3,4,5],5)