allowing randomization to be 0 in solver

Author: jonathan-taylor

selectiveInference/R/funs.randomized.R

@@ -22,14 +22,18 @@ fit_randomized_lasso = function(X,
 
     noise_type = match.arg(noise_type)
 
-    if (noise_type == 'gaussian') {
-       D = Norm(mean=0, sd=noise_scale)
+    if (noise_scale > 0) {
+        if (noise_type == 'gaussian') {
+            D = Norm(mean=0, sd=noise_scale)
+        }
+        else if (noise_type == 'laplace') {
+            D = DExp(rate = 1 / noise_scale) # D is a Laplace distribution with rate = 1.
+        }
+        perturb_ = distr::r(D)(p)
+    } else {
+        perturb_ = rep(0, p)
     }
-    else if (noise_type == 'laplace') {
-       D = DExp(rate = 1 / noise_scale) # D is a Laplace distribution with rate = 1.
-    }
-    perturb_ = distr::r(D)(p)
-			
+
     lam = as.numeric(lam)
     if (length(lam) == 1) {
        lam = rep(lam, p)

tests/test_QP.R

@@ -7,7 +7,7 @@ X = matrix(rnorm(n * p), n, p)
 Y = rnorm(n)
 lam = 2
 
-soln1 = selectiveInference:::fit_randomized_lasso(X, Y, lam, 1.e-12, 0)$soln
+soln1 = selectiveInference:::fit_randomized_lasso(X, Y, lam, 0, 0)$soln
 G = glmnet(X, Y, intercept=FALSE, standardize=FALSE)
 soln2 = coef(G, s=1/n, exact=TRUE, x=X, y=Y)[-1]