context("LOR interface") A_data <- rbinom(10, 15, 0.3) * 100 / 15 B_data <- rbinom(12, 15, 0.2) * 100 / 15 test_that("LOR handles scales correctly.", { expect_warning(LOR(A_data, B_data, scale = "count")) expect_warning(LOR(A_data, B_data, scale = "rate")) expect_error(LOR(A_data + 100, B_data)) LOR_pct <- LOR(A_data, B_data, scale = "percentage") LOR_prop <- LOR(A_data / 100, B_data / 100, scale = "proportion") expect_is(LOR_pct, "data.frame") expect_equal(LOR_pct, LOR_prop) LOR_pct <- LOR(A_data, B_data, scale = "percentage", bias_correct = FALSE) LOR_prop <- LOR(A_data / 100, B_data / 100, scale = "proportion", bias_correct = FALSE) expect_equal(LOR_pct, LOR_prop) })