EIF estimator with linear calibration (suboptimal for discrete)

Fits \(g(\hat{Y}) = a + b\cdot\hat{Y}\) using linear regression. This is suboptimal when \(\hat{Y}\) is discrete but shows importance of proper calibration.

Usage

eif_linear_discrete_point_and_ci(Y_cal, Yhat_cal, Yhat_test, alpha = 0.1)

Arguments

Y_cal

Vector of true labels on the calibration set.

Yhat_cal

Vector of surrogate predictions on the calibration set.

Yhat_test

Vector of surrogate predictions on the test set.

alpha

Miscoverage level for the confidence interval.