## somewhat arbitray values library(Renext) set.seed(149) lambda1 <- 1.0 lambda2 <- 1 + 10*rexp(1) alpha <- 0.3 test <- rep(NA, 100) for (i in 1:300) { ## Choose more values in the difficult zones p.test <- c(0.1+0.8*sort(runif(50)), 0.9+0.05*sort(runif(50)), 0.95 + 0.05*sort(runif(50))) H.test <- - log(1-p.test) q.test <- qmixexp2(p.test, prob1 = alpha, rate1 = lambda1, rate2 = lambda2) ## the result should be equal to p.test p2.test <- pmixexp2(q.test, prob1 = alpha, rate1 = lambda1, rate2 = lambda2) test[i] <- max(abs(p2.test - p.test)) } stopifnot(max(abs(test)) < 1e-8)