tests/testthat/test-regressionlinearbayesian.R
In jasp-stats/jaspRegression: Regression Module for JASP
context("Bayesian Linear Regression")
# does not test
# - plots (will/ should be updated in the future)
test_that("Main tables results match", {
set.seed(1)
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$modelPrior <- "betaBinomial"
options$dependent <- "contNormal"
options$covariates <- "contGamma"
options$weights <- "facFifty"
options$modelTerms <- list(
list(components="contGamma", isNuisance=FALSE)
)
options$posteriorSummaryPlot <- TRUE
options$posteriorSummaryTable <- TRUE
options$descriptives <- TRUE
options$setSeed <- TRUE
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "test.csv", options)
table <- results[["results"]][["basreg"]][["collection"]][["basreg_modelComparisonTable"]][["data"]]
jaspTools::expect_equal_tables(
table,
list("Null model", 1, 4.74865017735093, 0.826046120541498, 0, 0.5,
"contGamma", 0.210586158729818, 0.210586158729818, 0.173953879458502,
1.55940279678024e-06, 0.5),
label = "regressionTable"
)
table <- results[["results"]][["basreg"]][["collection"]][["basreg_postSumContainer"]][["collection"]][["basreg_postSumContainer_postSumTable"]][["data"]]
jaspTools::expect_equal_tables(
table,
list(1, "Intercept", -0.477863289393241, -0.255843391953333, 0, 0,
1, 1, 0.0989748100578513, -0.0887658028069459, 0.210586158729818,
"contGamma", -0.0481976534136304, -0.000116767975422579, 0.826046120541498,
0.5, 0.173953879458502, 0.5, 0.0241168865879483, 0.070739904433129
),
label = "posteriorSummaryTable"
)
table <- results[["results"]][["descriptivesTable"]][["data"]]
jaspTools::expect_equal_tables(
table,
list("contNormal", 100, -0.18874858754, 1.05841360919316, "contGamma",
100, 2.03296079621, 1.53241112621044),
label = "descriptivesTable"
)
plot <- results[["state"]][["figures"]][[1]][["obj"]]
jaspTools::expect_equal_plots(plot, "posteriorCoefficientsWithCRI", "RegressionLinearBayesian")
ybreaks <- jaspGraphs::getAxisBreaks(plot)[["y"]]
testthat::expect_true(
# to test if the ybreaks contain scientific notation, we print the ybreaks and check if there is any ... e-... in there
!any(grepl("e-", capture.output(print(ybreaks)), fixed = TRUE)),
label = "no scientific notation on the y-axis of the posterior coefficients plot"
)
})
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$covariates <- c("adverts", "airplay", "attract")
options$dependent <- "sales"
options$modelPrior <- "betaBinomial"
options$modelTerms <- list(list(components = "adverts", isNuisance = FALSE),
list(components = "airplay", isNuisance = FALSE),
list(components = c("adverts", "airplay"), isNuisance = FALSE))
options$posteriorSummaryTable <- TRUE
options$setSeed <- TRUE
set.seed(1)
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "Album Sales.csv", options)
test_that("Model Comparison - sales table results match", {
table <- results[["results"]][["basreg"]][["collection"]][["basreg_modelComparisonTable"]][["data"]]
jaspTools::expect_equal_tables(table,
list(1, 14.3947181553409, "adverts + airplay", 0.629285746598297, 0.642772909910809,
0.111111111111111, 0.185253135065882, 1.11151881039529, "adverts + airplay + adverts<unicode><unicode><unicode>airplay",
0.632270344815323, 0.357227090089191, 0.333333333333333, 8.35835904551863e-23,
4.29802141261388e-22, "airplay", 0.358703726117256, 5.37252676576735e-23,
0.111111111111111, 2.33573121426008e-24, 1.20107579948757e-23,
"adverts", 0.334648067623073, 1.50134474935946e-24, 0.111111111111111,
1.76804178544216e-40, 6.81869617963531e-40, "Null model", 0,
3.40934808981766e-40, 0.333333333333333),
label = "regressionTable")
})
test_that("Posterior Summaries of Coefficients table (all models) results match", {
table <- results[["results"]][["basreg"]][["collection"]][["basreg_postSumContainer"]][["collection"]][["basreg_postSumContainer_postSumTable"]][["data"]]
jaspTools::expect_equal_tables(table,
list(1, "Intercept", 185.584190973766, 193.2, 0, 0, 1, 1, 3.49006833453854,
199.210562797446, 1.48905726277148e+22, "adverts", 0.0699687329978213,
0.0957495623188114, 0, 0.444444444444444, 1, 0.555555555555556,
0.0196173572277313, 0.144714226721958, 5.32855628489937e+23,
"airplay", 2.93147903286366, 3.70853874547101, 0, 0.444444444444444,
1, 0.555555555555556, 0.405127356180258, 4.57041250208677, 1.11151881039529,
"adverts<unicode><unicode><unicode><unicode><unicode><unicode><unicode><unicode><unicode>airplay",
-0.00189399793265812, -0.000322869356058448, 0.642772909910809,
0.666666666666667, 0.357227090089191, 0.333333333333333, 0.000610819394600897,
0.000241181445566148))
})
options$effectsType <- "matchedModels"
set.seed(1)
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "Album Sales.csv", options)
test_that("Posterior Summaries of Coefficients table (matched models) results match", {
table <- results[["results"]][["basreg"]][["collection"]][["basreg_postSumContainer"]][["collection"]][["basreg_postSumContainer_postSumTable"]][["data"]]
jaspTools::expect_equal_tables(table,
list(1, "Intercept", 185.584190973766, 193.2, 0, 0, 1, 1, 3.49006833453854,
199.210562797446, 2.39281417453861e+22, "adverts", 0.0699687329978213,
0.0957495623188114, 5.37252676576734e-23, 0.444444444444444,
0.642772909910809, 0.222222222222222, 0.0196173572277313, 0.144714226721958,
8.56262907217075e+23, "airplay", 2.93147903286366, 3.70853874547101,
1.50134474935946e-24, 0.444444444444444, 0.642772909910809,
0.222222222222222, 0.405127356180258, 4.57041250208677, 0.185253135065882,
"adverts<unicode><unicode><unicode><unicode><unicode><unicode><unicode><unicode><unicode>airplay",
-0.00189399793265812, -0.000322869356058448, 0.642772909910809,
0.111111111111111, 0.357227090089191, 0.333333333333333, 0.000610819394600897,
0.000241181445566148))
})
test_that("Coefficient plots match", {
set.seed(1)
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$modelPrior <- "betaBinomial"
options$dependent <- "contNormal"
options$covariates <- list("contGamma", "debCollin1", "contcor2")
options$modelTerms <- list(
list(components="contGamma", isNuisance=FALSE),
list(components="debCollin1", isNuisance=FALSE),
list(components="contcor2", isNuisance=FALSE)
)
options$inclusionProbabilitiesPlot <- TRUE
options$marginalPosteriorPlot <- TRUE
options$posteriorSummaryPlotWithoutIntercept <- FALSE
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "test.csv", options)
inclusionProbabilities <- results[['state']][['figures']][[1]][["obj"]]
jaspTools::expect_equal_plots(inclusionProbabilities, "inclusionProbabilities", "RegressionLinearBayesian")
posteriorCoefficients <- results[['state']][['figures']][[2]][["obj"]]
jaspTools::expect_equal_plots(posteriorCoefficients, "posteriorCoefficients", "RegressionLinearBayesian")
})
test_that("Residuals plots match", {
set.seed(1)
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$modelPrior <- "betaBinomial"
options$dependent <- "contNormal"
options$covariates <- list("contGamma")
options$modelTerms <- list(
list(components="contGamma", isNuisance=FALSE)
)
options$residualsVsFittedPlot <- TRUE
options$qqPlot <- TRUE
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "test.csv", options)
residualsVsFitted <- results[['state']][['figures']][[1]][["obj"]]
jaspTools::expect_equal_plots(residualsVsFitted, "residualsVsFitted", "RegressionLinearBayesian")
qqPlot <- results[['state']][['figures']][[2]][["obj"]]
jaspTools::expect_equal_plots(qqPlot, "qqPlot", "RegressionLinearBayesian")
})
test_that("Models plots match", {
set.seed(1)
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$modelPrior <- "betaBinomial"
options$dependent <- "contNormal"
options$covariates <- list("contGamma", "contExpon", "contcor1")
options$modelTerms <- list(
list(components="contGamma", isNuisance=FALSE),
list(components="contExpon", isNuisance=FALSE),
list(components="contcor1", isNuisance=FALSE)
)
options$logPosteriorOddsPlot <- TRUE
options$modelComplexityPlot <- TRUE
options$modelProbabilitiesPlot <- TRUE
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "test.csv", options)
logPosteriorOdds <- results[['state']][['figures']][[1]][["obj"]]
jaspTools::expect_equal_plots(logPosteriorOdds, "logPosteriorOdds", "RegressionLinearBayesian")
modelProbabilities <- results[['state']][['figures']][[2]][["obj"]]
jaspTools::expect_equal_plots(modelProbabilities, "modelProbabilities", "RegressionLinearBayesian")
modelComplexity <- results[['state']][['figures']][[3]][["obj"]]
jaspTools::expect_equal_plots(modelComplexity, "modelComplexity", "RegressionLinearBayesian")
})
test_that("Model priors match", {
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$dependent <- "contNormal"
options$covariates <- list("contGamma", "contExpon", "contcor1")
options$modelTerms <- list(
list(components="contGamma", isNuisance=FALSE),
list(components="contExpon", isNuisance=FALSE),
list(components="contcor1", isNuisance=FALSE)
)
modelPriors <- list(
uniform = list(modelPrior = "uniform"),
betabinomial = list(modelPrior = "betaBinomial", betaBinomialParamA = 2, betaBinomialParamB = 3),
wilson = list(modelPrior = "wilson", wilsonParamLambda = 2),
bernoulli = list(modelPrior = "bernoulli", bernoulliParam = 0.75),
castillo = list(modelPrior = "castillo", castilloParamU = 2)
)
tables <- list(
uniform = list(1, 3.90143837337827, "Null model", 0, 0.357882899462674, 0.125,
0.667063040584368, 2.19516621516207, "contcor1", 0.0259312818065142,
0.238730455088721, 0.125, 0.250266509028401, 0.688641679428745,
"contExpon", 0.00394490646287682, 0.0895661038894857, 0.125,
0.245477808280897, 0.674195829319842, "contGamma", 0.00350468569669571,
0.0878523097813097, 0.125, 0.24073279559456, 0.659935214199163,
"contExpon + contcor1", 0.0305485694813512, 0.0861541508831363,
0.125, 0.207447147529827, 0.561369587657518, "contGamma + contcor1",
0.0271156286160025, 0.0742417866432354, 0.125, 0.0916338337573983,
0.237342735695682, "contGamma + contExpon", 0.00794701370052642,
0.0327941821139783, 0.125, 0.0915889308672561, 0.237222490352522,
"contGamma + contExpon + contcor1", 0.0320272344954695, 0.03277811213746,
0.125),
betabinomial = list(1, 3.76674705968188, "Null model", 0, 0.601068947543112, 0.285714285714286,
0.667063040584368, 1.48036998661697, "contcor1", 0.0259312818065142,
0.160380351899582, 0.114285714285714, 0.250266509028401, 0.496180698646568,
"contExpon", 0.00394490646287682, 0.060170970874796, 0.114285714285714,
0.245477808280897, 0.486091091246406, "contGamma", 0.00350468569669571,
0.0590196351474355, 0.114285714285714, 0.24073279559456, 0.48404226619047,
"contExpon + contcor1", 0.0305485694813512, 0.04340910242614,
0.0857142857142857, 0.207447147529827, 0.414513847951994, "contGamma + contcor1",
0.0271156286160025, 0.0374070115909721, 0.0857142857142857,
0.0915889308672561, 0.17450152476059, "contGamma + contExpon + contcor1",
0.0320272344954695, 0.0220205049131922, 0.114285714285714, 0.0916338337573983,
0.179211605034763, "contGamma + contExpon", 0.00794701370052642,
0.01652347560477, 0.0857142857142857),
wilson = list(1, 3.00723842615036, "Null model", 0, 0.857437693351001, 0.666666666666667,
0.667063040584368, 0.847009328691918, "contcor1", 0.0259312818065142,
0.0714956243547957, 0.0833333333333333, 0.250266509028401, 0.303191058063168,
"contExpon", 0.00394490646287682, 0.02682349227804, 0.0833333333333333,
0.245477808280897, 0.297232917444893, "contGamma", 0.00350468569669571,
0.0263102407126539, 0.0833333333333333, 0.24073279559456, 0.304492845967535,
"contExpon + contcor1", 0.0305485694813512, 0.00737190617744777,
0.0238095238095238, 0.207447147529827, 0.262122056715543, "contGamma + contcor1",
0.0271156286160025, 0.00635260727393641, 0.0238095238095238,
0.0916338337573983, 0.115373118783951, "contGamma + contExpon",
0.00794701370052642, 0.00280608225178045, 0.0238095238095238,
0.0915889308672561, 0.116558805489141, "contGamma + contExpon + contcor1",
0.0320272344954695, 0.00140235360034472, 0.0119047619047619),
bernoulli = list(1, 0.362563657801133, "contGamma + contExpon + contcor1", 0.0320272344954695,
0.209219538601646, 0.421875, 2.62840491001543, 1.37161839640395,
"contExpon + contcor1", 0.0305485694813512, 0.183304554177243,
0.140625, 7.28322772487837, 4.14432135426702, "contcor1", 0.0259312818065142,
0.169310393792196, 0.046875, 2.26498055567969, 1.14639058082694,
"contGamma + contcor1", 0.0271156286160025, 0.157959395600335,
0.140625, 10.9183499635927, 5.82274069025381, "Null model",
0, 0.08460489413608, 0.015625, 1.00049026546895, 0.458379911234064,
"contGamma + contExpon", 0.00794701370052642, 0.0697740372389506,
0.140625, 2.73249732973872, 1.37920924591115, "contExpon", 0.00394490646287682,
0.0635213145064626, 0.046875, 2.68021261910655, 1.35106536878368,
"contGamma", 0.00350468569669571, 0.0623058719470867, 0.046875),
castillo = list(1, 2.87159171765746, "Null model", 0, 0.895993407636698, 0.75,
0.667063040584368, 0.785243329949621, "contcor1", 0.0259312818065142,
0.0543349169856078, 0.0681818181818182, 0.250266509028401, 0.284395111302656,
"contExpon", 0.00394490646287682, 0.0203851947492453, 0.0681818181818182,
0.245477808280897, 0.278842351875457, "contGamma", 0.00350468569669571,
0.0199951361764354, 0.0681818181818182, 0.24073279559456, 0.284788465921838,
"contExpon + contcor1", 0.0305485694813512, 0.00392172723372142,
0.0136363636363636, 0.207447147529827, 0.24527780194552, "contGamma + contcor1",
0.0271156286160025, 0.00337947775672295, 0.0136363636363636,
0.0916338337573983, 0.108139723436462, "contGamma + contExpon",
0.00794701370052642, 0.00149278747205647, 0.0136363636363636,
0.0915889308672561, 0.10897428428032, "contGamma + contExpon + contcor1",
0.0320272344954695, 0.000497351989512453, 0.00454545454545455)
)
for (nm in names(modelPriors)) {
set.seed(1)
options[names(modelPriors[[nm]])] <- modelPriors[[nm]]
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "test.csv", options)
table <- results[["results"]][["basreg"]][["collection"]][["basreg_modelComparisonTable"]][["data"]]
jaspTools::expect_equal_tables(
table,
tables[[nm]],
label = paste("regressionTable modelprior:", paste(modelPriors[[nm]], collapse = ""))
)
}
})
test_that("Exporting residuals works", {
data("Hald", package = "BAS")
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$dependent <- "Y"
options$covariates <- paste0("X", 1:4)
options$modelTerms <- lapply(options$covariates, function(x) list(components = x, isNuisance = FALSE))
options$modelPrior <- "betaBinomial"
options$priorRegressionCoefficients <- "gPrior"
options$gPriorAlpha <- 13
options$residualsSavedToData <- TRUE
options$residualSdsSavedToData <- TRUE
summaryTypes <- c("best", "complex", "median", "averaged")
for (summaryType in summaryTypes) {
options$summaryType <- summaryType
options$residualsSavedToDataColumn <- paste("residualsSavedToData", "-", summaryType)
options$residualSdsSavedToDataColumn <- paste("residualSdsSavedToData", "-", summaryType)
results <- jaspTools::runAnalysis("RegressionLinearBayesian", Hald, options)
# ideally we'd check the actual data here but the jaspTools output unfortunately does not contain the data
testthat::expect_identical(
results[["results"]][["basreg"]][["collection"]][["basreg_residualsSavedToDataColumn"]][c("columnName", "columnType")],
list(columnName = options$residualsSavedToDataColumn, columnType = "scale")
)
testthat::expect_identical(
results[["results"]][["basreg"]][["collection"]][["basreg_residualSdsSavedToDataColumn"]][c("columnName", "columnType")],
list(columnName = options$residualSdsSavedToDataColumn, columnType = "scale")
)
}
})
jasp-stats/jaspRegression documentation built on May 5, 2024, 10:55 p.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
context("Bayesian Linear Regression")
# does not test
# - plots (will/ should be updated in the future)
test_that("Main tables results match", {
set.seed(1)
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$modelPrior <- "betaBinomial"
options$dependent <- "contNormal"
options$covariates <- "contGamma"
options$weights <- "facFifty"
options$modelTerms <- list(
list(components="contGamma", isNuisance=FALSE)
)
options$posteriorSummaryPlot <- TRUE
options$posteriorSummaryTable <- TRUE
options$descriptives <- TRUE
options$setSeed <- TRUE
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "test.csv", options)
table <- results[["results"]][["basreg"]][["collection"]][["basreg_modelComparisonTable"]][["data"]]
jaspTools::expect_equal_tables(
table,
list("Null model", 1, 4.74865017735093, 0.826046120541498, 0, 0.5,
"contGamma", 0.210586158729818, 0.210586158729818, 0.173953879458502,
1.55940279678024e-06, 0.5),
label = "regressionTable"
)
table <- results[["results"]][["basreg"]][["collection"]][["basreg_postSumContainer"]][["collection"]][["basreg_postSumContainer_postSumTable"]][["data"]]
jaspTools::expect_equal_tables(
table,
list(1, "Intercept", -0.477863289393241, -0.255843391953333, 0, 0,
1, 1, 0.0989748100578513, -0.0887658028069459, 0.210586158729818,
"contGamma", -0.0481976534136304, -0.000116767975422579, 0.826046120541498,
0.5, 0.173953879458502, 0.5, 0.0241168865879483, 0.070739904433129
),
label = "posteriorSummaryTable"
)
table <- results[["results"]][["descriptivesTable"]][["data"]]
jaspTools::expect_equal_tables(
table,
list("contNormal", 100, -0.18874858754, 1.05841360919316, "contGamma",
100, 2.03296079621, 1.53241112621044),
label = "descriptivesTable"
)
plot <- results[["state"]][["figures"]][[1]][["obj"]]
jaspTools::expect_equal_plots(plot, "posteriorCoefficientsWithCRI", "RegressionLinearBayesian")
ybreaks <- jaspGraphs::getAxisBreaks(plot)[["y"]]
testthat::expect_true(
# to test if the ybreaks contain scientific notation, we print the ybreaks and check if there is any ... e-... in there
!any(grepl("e-", capture.output(print(ybreaks)), fixed = TRUE)),
label = "no scientific notation on the y-axis of the posterior coefficients plot"
)
})
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$covariates <- c("adverts", "airplay", "attract")
options$dependent <- "sales"
options$modelPrior <- "betaBinomial"
options$modelTerms <- list(list(components = "adverts", isNuisance = FALSE),
list(components = "airplay", isNuisance = FALSE),
list(components = c("adverts", "airplay"), isNuisance = FALSE))
options$posteriorSummaryTable <- TRUE
options$setSeed <- TRUE
set.seed(1)
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "Album Sales.csv", options)
test_that("Model Comparison - sales table results match", {
table <- results[["results"]][["basreg"]][["collection"]][["basreg_modelComparisonTable"]][["data"]]
jaspTools::expect_equal_tables(table,
list(1, 14.3947181553409, "adverts + airplay", 0.629285746598297, 0.642772909910809,
0.111111111111111, 0.185253135065882, 1.11151881039529, "adverts + airplay + adverts<unicode><unicode><unicode>airplay",
0.632270344815323, 0.357227090089191, 0.333333333333333, 8.35835904551863e-23,
4.29802141261388e-22, "airplay", 0.358703726117256, 5.37252676576735e-23,
0.111111111111111, 2.33573121426008e-24, 1.20107579948757e-23,
"adverts", 0.334648067623073, 1.50134474935946e-24, 0.111111111111111,
1.76804178544216e-40, 6.81869617963531e-40, "Null model", 0,
3.40934808981766e-40, 0.333333333333333),
label = "regressionTable")
})
test_that("Posterior Summaries of Coefficients table (all models) results match", {
table <- results[["results"]][["basreg"]][["collection"]][["basreg_postSumContainer"]][["collection"]][["basreg_postSumContainer_postSumTable"]][["data"]]
jaspTools::expect_equal_tables(table,
list(1, "Intercept", 185.584190973766, 193.2, 0, 0, 1, 1, 3.49006833453854,
199.210562797446, 1.48905726277148e+22, "adverts", 0.0699687329978213,
0.0957495623188114, 0, 0.444444444444444, 1, 0.555555555555556,
0.0196173572277313, 0.144714226721958, 5.32855628489937e+23,
"airplay", 2.93147903286366, 3.70853874547101, 0, 0.444444444444444,
1, 0.555555555555556, 0.405127356180258, 4.57041250208677, 1.11151881039529,
"adverts<unicode><unicode><unicode><unicode><unicode><unicode><unicode><unicode><unicode>airplay",
-0.00189399793265812, -0.000322869356058448, 0.642772909910809,
0.666666666666667, 0.357227090089191, 0.333333333333333, 0.000610819394600897,
0.000241181445566148))
})
options$effectsType <- "matchedModels"
set.seed(1)
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "Album Sales.csv", options)
test_that("Posterior Summaries of Coefficients table (matched models) results match", {
table <- results[["results"]][["basreg"]][["collection"]][["basreg_postSumContainer"]][["collection"]][["basreg_postSumContainer_postSumTable"]][["data"]]
jaspTools::expect_equal_tables(table,
list(1, "Intercept", 185.584190973766, 193.2, 0, 0, 1, 1, 3.49006833453854,
199.210562797446, 2.39281417453861e+22, "adverts", 0.0699687329978213,
0.0957495623188114, 5.37252676576734e-23, 0.444444444444444,
0.642772909910809, 0.222222222222222, 0.0196173572277313, 0.144714226721958,
8.56262907217075e+23, "airplay", 2.93147903286366, 3.70853874547101,
1.50134474935946e-24, 0.444444444444444, 0.642772909910809,
0.222222222222222, 0.405127356180258, 4.57041250208677, 0.185253135065882,
"adverts<unicode><unicode><unicode><unicode><unicode><unicode><unicode><unicode><unicode>airplay",
-0.00189399793265812, -0.000322869356058448, 0.642772909910809,
0.111111111111111, 0.357227090089191, 0.333333333333333, 0.000610819394600897,
0.000241181445566148))
})
test_that("Coefficient plots match", {
set.seed(1)
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$modelPrior <- "betaBinomial"
options$dependent <- "contNormal"
options$covariates <- list("contGamma", "debCollin1", "contcor2")
options$modelTerms <- list(
list(components="contGamma", isNuisance=FALSE),
list(components="debCollin1", isNuisance=FALSE),
list(components="contcor2", isNuisance=FALSE)
)
options$inclusionProbabilitiesPlot <- TRUE
options$marginalPosteriorPlot <- TRUE
options$posteriorSummaryPlotWithoutIntercept <- FALSE
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "test.csv", options)
inclusionProbabilities <- results[['state']][['figures']][[1]][["obj"]]
jaspTools::expect_equal_plots(inclusionProbabilities, "inclusionProbabilities", "RegressionLinearBayesian")
posteriorCoefficients <- results[['state']][['figures']][[2]][["obj"]]
jaspTools::expect_equal_plots(posteriorCoefficients, "posteriorCoefficients", "RegressionLinearBayesian")
})
test_that("Residuals plots match", {
set.seed(1)
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$modelPrior <- "betaBinomial"
options$dependent <- "contNormal"
options$covariates <- list("contGamma")
options$modelTerms <- list(
list(components="contGamma", isNuisance=FALSE)
)
options$residualsVsFittedPlot <- TRUE
options$qqPlot <- TRUE
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "test.csv", options)
residualsVsFitted <- results[['state']][['figures']][[1]][["obj"]]
jaspTools::expect_equal_plots(residualsVsFitted, "residualsVsFitted", "RegressionLinearBayesian")
qqPlot <- results[['state']][['figures']][[2]][["obj"]]
jaspTools::expect_equal_plots(qqPlot, "qqPlot", "RegressionLinearBayesian")
})
test_that("Models plots match", {
set.seed(1)
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$modelPrior <- "betaBinomial"
options$dependent <- "contNormal"
options$covariates <- list("contGamma", "contExpon", "contcor1")
options$modelTerms <- list(
list(components="contGamma", isNuisance=FALSE),
list(components="contExpon", isNuisance=FALSE),
list(components="contcor1", isNuisance=FALSE)
)
options$logPosteriorOddsPlot <- TRUE
options$modelComplexityPlot <- TRUE
options$modelProbabilitiesPlot <- TRUE
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "test.csv", options)
logPosteriorOdds <- results[['state']][['figures']][[1]][["obj"]]
jaspTools::expect_equal_plots(logPosteriorOdds, "logPosteriorOdds", "RegressionLinearBayesian")
modelProbabilities <- results[['state']][['figures']][[2]][["obj"]]
jaspTools::expect_equal_plots(modelProbabilities, "modelProbabilities", "RegressionLinearBayesian")
modelComplexity <- results[['state']][['figures']][[3]][["obj"]]
jaspTools::expect_equal_plots(modelComplexity, "modelComplexity", "RegressionLinearBayesian")
})
test_that("Model priors match", {
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$dependent <- "contNormal"
options$covariates <- list("contGamma", "contExpon", "contcor1")
options$modelTerms <- list(
list(components="contGamma", isNuisance=FALSE),
list(components="contExpon", isNuisance=FALSE),
list(components="contcor1", isNuisance=FALSE)
)
modelPriors <- list(
uniform = list(modelPrior = "uniform"),
betabinomial = list(modelPrior = "betaBinomial", betaBinomialParamA = 2, betaBinomialParamB = 3),
wilson = list(modelPrior = "wilson", wilsonParamLambda = 2),
bernoulli = list(modelPrior = "bernoulli", bernoulliParam = 0.75),
castillo = list(modelPrior = "castillo", castilloParamU = 2)
)
tables <- list(
uniform = list(1, 3.90143837337827, "Null model", 0, 0.357882899462674, 0.125,
0.667063040584368, 2.19516621516207, "contcor1", 0.0259312818065142,
0.238730455088721, 0.125, 0.250266509028401, 0.688641679428745,
"contExpon", 0.00394490646287682, 0.0895661038894857, 0.125,
0.245477808280897, 0.674195829319842, "contGamma", 0.00350468569669571,
0.0878523097813097, 0.125, 0.24073279559456, 0.659935214199163,
"contExpon + contcor1", 0.0305485694813512, 0.0861541508831363,
0.125, 0.207447147529827, 0.561369587657518, "contGamma + contcor1",
0.0271156286160025, 0.0742417866432354, 0.125, 0.0916338337573983,
0.237342735695682, "contGamma + contExpon", 0.00794701370052642,
0.0327941821139783, 0.125, 0.0915889308672561, 0.237222490352522,
"contGamma + contExpon + contcor1", 0.0320272344954695, 0.03277811213746,
0.125),
betabinomial = list(1, 3.76674705968188, "Null model", 0, 0.601068947543112, 0.285714285714286,
0.667063040584368, 1.48036998661697, "contcor1", 0.0259312818065142,
0.160380351899582, 0.114285714285714, 0.250266509028401, 0.496180698646568,
"contExpon", 0.00394490646287682, 0.060170970874796, 0.114285714285714,
0.245477808280897, 0.486091091246406, "contGamma", 0.00350468569669571,
0.0590196351474355, 0.114285714285714, 0.24073279559456, 0.48404226619047,
"contExpon + contcor1", 0.0305485694813512, 0.04340910242614,
0.0857142857142857, 0.207447147529827, 0.414513847951994, "contGamma + contcor1",
0.0271156286160025, 0.0374070115909721, 0.0857142857142857,
0.0915889308672561, 0.17450152476059, "contGamma + contExpon + contcor1",
0.0320272344954695, 0.0220205049131922, 0.114285714285714, 0.0916338337573983,
0.179211605034763, "contGamma + contExpon", 0.00794701370052642,
0.01652347560477, 0.0857142857142857),
wilson = list(1, 3.00723842615036, "Null model", 0, 0.857437693351001, 0.666666666666667,
0.667063040584368, 0.847009328691918, "contcor1", 0.0259312818065142,
0.0714956243547957, 0.0833333333333333, 0.250266509028401, 0.303191058063168,
"contExpon", 0.00394490646287682, 0.02682349227804, 0.0833333333333333,
0.245477808280897, 0.297232917444893, "contGamma", 0.00350468569669571,
0.0263102407126539, 0.0833333333333333, 0.24073279559456, 0.304492845967535,
"contExpon + contcor1", 0.0305485694813512, 0.00737190617744777,
0.0238095238095238, 0.207447147529827, 0.262122056715543, "contGamma + contcor1",
0.0271156286160025, 0.00635260727393641, 0.0238095238095238,
0.0916338337573983, 0.115373118783951, "contGamma + contExpon",
0.00794701370052642, 0.00280608225178045, 0.0238095238095238,
0.0915889308672561, 0.116558805489141, "contGamma + contExpon + contcor1",
0.0320272344954695, 0.00140235360034472, 0.0119047619047619),
bernoulli = list(1, 0.362563657801133, "contGamma + contExpon + contcor1", 0.0320272344954695,
0.209219538601646, 0.421875, 2.62840491001543, 1.37161839640395,
"contExpon + contcor1", 0.0305485694813512, 0.183304554177243,
0.140625, 7.28322772487837, 4.14432135426702, "contcor1", 0.0259312818065142,
0.169310393792196, 0.046875, 2.26498055567969, 1.14639058082694,
"contGamma + contcor1", 0.0271156286160025, 0.157959395600335,
0.140625, 10.9183499635927, 5.82274069025381, "Null model",
0, 0.08460489413608, 0.015625, 1.00049026546895, 0.458379911234064,
"contGamma + contExpon", 0.00794701370052642, 0.0697740372389506,
0.140625, 2.73249732973872, 1.37920924591115, "contExpon", 0.00394490646287682,
0.0635213145064626, 0.046875, 2.68021261910655, 1.35106536878368,
"contGamma", 0.00350468569669571, 0.0623058719470867, 0.046875),
castillo = list(1, 2.87159171765746, "Null model", 0, 0.895993407636698, 0.75,
0.667063040584368, 0.785243329949621, "contcor1", 0.0259312818065142,
0.0543349169856078, 0.0681818181818182, 0.250266509028401, 0.284395111302656,
"contExpon", 0.00394490646287682, 0.0203851947492453, 0.0681818181818182,
0.245477808280897, 0.278842351875457, "contGamma", 0.00350468569669571,
0.0199951361764354, 0.0681818181818182, 0.24073279559456, 0.284788465921838,
"contExpon + contcor1", 0.0305485694813512, 0.00392172723372142,
0.0136363636363636, 0.207447147529827, 0.24527780194552, "contGamma + contcor1",
0.0271156286160025, 0.00337947775672295, 0.0136363636363636,
0.0916338337573983, 0.108139723436462, "contGamma + contExpon",
0.00794701370052642, 0.00149278747205647, 0.0136363636363636,
0.0915889308672561, 0.10897428428032, "contGamma + contExpon + contcor1",
0.0320272344954695, 0.000497351989512453, 0.00454545454545455)
)
for (nm in names(modelPriors)) {
set.seed(1)
options[names(modelPriors[[nm]])] <- modelPriors[[nm]]
results <- jaspTools::runAnalysis("RegressionLinearBayesian", "test.csv", options)
table <- results[["results"]][["basreg"]][["collection"]][["basreg_modelComparisonTable"]][["data"]]
jaspTools::expect_equal_tables(
table,
tables[[nm]],
label = paste("regressionTable modelprior:", paste(modelPriors[[nm]], collapse = ""))
)
}
})
test_that("Exporting residuals works", {
data("Hald", package = "BAS")
options <- jaspTools::analysisOptions("RegressionLinearBayesian")
options$dependent <- "Y"
options$covariates <- paste0("X", 1:4)
options$modelTerms <- lapply(options$covariates, function(x) list(components = x, isNuisance = FALSE))
options$modelPrior <- "betaBinomial"
options$priorRegressionCoefficients <- "gPrior"
options$gPriorAlpha <- 13
options$residualsSavedToData <- TRUE
options$residualSdsSavedToData <- TRUE
summaryTypes <- c("best", "complex", "median", "averaged")
for (summaryType in summaryTypes) {
options$summaryType <- summaryType
options$residualsSavedToDataColumn <- paste("residualsSavedToData", "-", summaryType)
options$residualSdsSavedToDataColumn <- paste("residualSdsSavedToData", "-", summaryType)
results <- jaspTools::runAnalysis("RegressionLinearBayesian", Hald, options)
# ideally we'd check the actual data here but the jaspTools output unfortunately does not contain the data
testthat::expect_identical(
results[["results"]][["basreg"]][["collection"]][["basreg_residualsSavedToDataColumn"]][c("columnName", "columnType")],
list(columnName = options$residualsSavedToDataColumn, columnType = "scale")
)
testthat::expect_identical(
results[["results"]][["basreg"]][["collection"]][["basreg_residualSdsSavedToDataColumn"]][c("columnName", "columnType")],
list(columnName = options$residualSdsSavedToDataColumn, columnType = "scale")
)
}
})
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.