#------------------------------------------------------------------------------- # Copyright (c) 2019-2022 University of Newcastle upon Tyne. All rights reserved. # # This program and the accompanying materials # are made available under the terms of the GNU Public License v3.0. # # You should have received a copy of the GNU General Public License # along with this program. If not, see . #------------------------------------------------------------------------------- # # Set up # context("ds.elspline::smk::setup") connect.studies.dataset.cnsim(list("LAB_TRIG", "PM_BMI_CONTINUOUS")) test_that("setup", { ds_expect_variables(c("D")) }) # # Tests # context("ds.elspline::smk::test1") test_that("elspline", { ds.elspline(x="D$PM_BMI_CONTINUOUS", n=3, newobj="elsplineDS", datasources=ds.test_env$connections) res.class <- ds.class("elsplineDS", datasources=ds.test_env$connections) expect_length(res.class, 3) expect_equal(res.class$sim1[1], "lspline") expect_equal(res.class$sim1[2], "matrix") expect_equal(res.class$sim2[1], "lspline") expect_equal(res.class$sim2[2], "matrix") expect_equal(res.class$sim3[1], "lspline") expect_equal(res.class$sim3[2], "matrix") res.mod <- ds.glm(formula = "D$LAB_TRIG~elsplineDS", family='gaussian', datasources=ds.test_env$connections) expect_length(res.mod, 13) expect_equal(res.mod$Nvalid, 7477) expect_equal(res.mod$Nmissing, 1902) expect_equal(res.mod$Ntotal, 9379) expect_length(res.mod$disclosure.risk, 3) expect_equal(res.mod$disclosure[1], 0) expect_equal(res.mod$disclosure[3], 0) expect_equal(res.mod$disclosure[2], 0) expect_length(res.mod$errorMessage, 3) expect_equal(res.mod$errorMessage[1], "No errors") expect_equal(res.mod$errorMessage[2], "No errors") expect_equal(res.mod$errorMessage[3], "No errors") expect_equal(res.mod$nsubs, 7477) expect_equal(res.mod$iter, 3) expect_true("family" %in% class(res.mod$family)) expect_equal(res.mod$formula, "D$LAB_TRIG ~ elsplineDS") expect_true("matrix" %in% class(res.mod$coefficients)) expect_true("array" %in% class(res.mod$coefficients)) expect_equal(res.mod$coefficients['(Intercept)','Estimate'], -0.35935028, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS1','Estimate'], 0.09176889, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS2','Estimate'], 0.06794047, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS3','Estimate'], 0.21256245, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['(Intercept)','Std. Error'], 0.144607412, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS1','Std. Error'], 0.006245815, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS2','Std. Error'], 0.004274001, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS3','Std. Error'], 0.021359728, tolerance = ds.test_env$tolerance) expect_equal(res.mod$dev, 16981.52, tolerance = ds.test_env$tolerance) expect_equal(res.mod$df, 7473) expect_equal(res.mod$output.information, "SEE TOP OF OUTPUT FOR INFORMATION ON MISSING DATA AND ERROR MESSAGES") }) context("ds.elspline::smk::test2") test_that("elspline", { ds.elspline(x="D$PM_BMI_CONTINUOUS", n=5, newobj="elsplineDS2", datasources=ds.test_env$connections) res.class <- ds.class("elsplineDS2", datasources=ds.test_env$connections) expect_length(res.class, 3) expect_equal(res.class$sim1[1], "lspline") expect_equal(res.class$sim1[2], "matrix") expect_equal(res.class$sim2[1], "lspline") expect_equal(res.class$sim2[2], "matrix") expect_equal(res.class$sim3[1], "lspline") expect_equal(res.class$sim3[2], "matrix") res.mod <- ds.glm(formula = "D$LAB_TRIG~elsplineDS2", family='gaussian', datasources=ds.test_env$connections) expect_length(res.mod, 13) expect_equal(res.mod$Nvalid, 7477) expect_equal(res.mod$Nmissing, 1902) expect_equal(res.mod$Ntotal, 9379) expect_length(res.mod$disclosure.risk, 3) expect_equal(res.mod$disclosure[1], 0) expect_equal(res.mod$disclosure[3], 0) expect_equal(res.mod$disclosure[2], 0) expect_length(res.mod$errorMessage, 3) expect_equal(res.mod$errorMessage[1], "No errors") expect_equal(res.mod$errorMessage[2], "No errors") expect_equal(res.mod$errorMessage[3], "No errors") expect_equal(res.mod$nsubs, 7477) expect_equal(res.mod$iter, 3) expect_true("family" %in% class(res.mod$family)) expect_equal(res.mod$formula, "D$LAB_TRIG ~ elsplineDS2") expect_true("matrix" %in% class(res.mod$coefficients)) expect_true("array" %in% class(res.mod$coefficients)) expect_equal(res.mod$coefficients['(Intercept)','Estimate'], -0.29808175, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS21','Estimate'], 0.08295113, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS22','Estimate'], 0.10199060, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS23','Estimate'], 0.05891621, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS24','Estimate'], 0.07761427, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS25','Estimate'], 0.65329224, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['(Intercept)','Std. Error'], 0.142080199, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS21','Std. Error'], 0.007062320, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS22','Std. Error'], 0.009361885, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS23','Std. Error'], 0.007798950, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS24','Std. Error'], 0.016254429, tolerance = ds.test_env$tolerance) expect_equal(res.mod$coefficients['elsplineDS25','Std. Error'], 0.067541719, tolerance = ds.test_env$tolerance) expect_equal(res.mod$dev, 16876.12, tolerance = ds.test_env$tolerance) expect_equal(res.mod$df, 7471) expect_equal(res.mod$output.information, "SEE TOP OF OUTPUT FOR INFORMATION ON MISSING DATA AND ERROR MESSAGES") }) # # Done # context("ds.elspline::smk::shutdown") test_that("shutdown", { ds_expect_variables(c("D", "elsplineDS", "elsplineDS2", "LAB_TRIG")) }) disconnect.studies.dataset.cnsim() context("ds.elspline::smk::done")