set.seed(1234) test_that("Toy GEE analysis creates correct output", { get_test4 = function(){ # To recreate input file from vignette 2 # write.csv(example2a$trial, file = "inst/extdata/exampleCRT.txt", row.names = FALSE) trial <- readdata("exampleCRT.txt") test_Estimates <- CRTanalysis(trial = trial, method = 'GEE',excludeBuffer = FALSE, alpha = 0.2) value <- round(test_Estimates$pt_ests$effect_size*1000) return(value) } expect_equal(get_test4(), 388) }) set.seed(1234) get_test5 = function(){ test_locationsLatLong <- readdata("example_latlong.csv") test_locationsxy <- latlong_as_xy(test_locationsLatLong) #test_site is simulated test_anonymizedlocations <- anonymize_site(test_locationsxy) test_clusters <- specify_clusters(test_anonymizedlocations,h = 5) test_arms <- randomizeCRT(trial = test_clusters,matchedPair = FALSE) test_buffer <- specify_buffer(trial = test_arms, buffer_width = 0.1) trial <- test_buffer$trial trial$cluster <- as.numeric(trial$cluster) trial$arm <- as.character(trial$arm) # To recreate test file # write.csv(trial, file = "inst/extdata/example_buffer.csv", row.names = FALSE) return(trial)} test_that("Anonymisation, randomization, and creation of buffer produces expected trial", { expect_equal(get_test5(), readdata("example_buffer.csv")) }) test_that("Analysis using T option gives expected efficacy", { get_test6 = function(extdata){ analysis <- CRTanalysis(readdata("exampleCRT.txt"),method = 'T',link='identity') value <- round(as.numeric(10000 * analysis$pt_ests$effect_size)) return(value)} expect_equal(get_test6(extdata), 1446) }) test_that("Analysis using GEE option gives expected coefficient of variation", { get_test7 = function(extdata){ analysis <- CRTanalysis(readdata("exampleCRT.txt"),method = 'GEE',link='log') value <- round(as.numeric(10 * analysis$description$cv_percent)) return(value)} expect_equal(get_test7(extdata), 484) })