tests/testthat/test-prediction.R

In dynConfiR: Dynamic Models for Confidence and Response Time Distributions

fits <- structure(list(model = c("dynWEV", "PCRMt", "dynWEV", "PCRMt"),
                       sbj = c(1, 1, 8, 8),
                       negLogLik = c(20.6368391020115, 31.2510706889092, 17.5490953740148, 26.335957708293),
                       N = c(17, 17, 17, 17),
                       k = c(23, 17, 23, 19),
                       BIC = c(106.437585117316, 110.666768226774, 100.262097661323, 106.502968953654),
                       AICc = c(-57.2977503674057, -447.497858622182, -63.473237823399, -137.328084583414), AIC = c(87.2736782040229,
                                                                                                                    96.5021413778184, 81.0981907480295, 90.671915416586),
                       t0 = c(0, 0, 0.319525056128088, 0.311195873979081),
                       st0 = c(1.12963012317701, 1.38876864436006, 0.530550462493294, 0.550715158444022),
                       v1 = c(0.172998967009845, 0.803225303592474, 0.200143716989775, 0.103200570174599),
                       v2 = c(0.241457085865768, 0.257772690122886, 0.340016451531057, 0.173222271497872),
                       v3 = c(0.676480108664119,       0.612117874357785, 1.24517397378776, 0.665734848622986),
                       v4 = c(1.5499032887235,      1.72173493421176, 2.82777200274886, 1.6556441561065),
                       v5 = c(2.53678081878952,   3.39752560799776, 5.01116887971942, 4.009703275535),
                       thetaLower1 = c(1.17622428639492,    0.17179837861488, 0.789706736239607, 0.879092614245285),
                       thetaLower2 = c(1.17622428639492,   0.17179837861488, 1.42924978336366, 1.45498757702495),
                       thetaLower3 = c(1.17622428639492,      1.179837861488, 2.06880417585998, 2.02835133872529),
                       thetaLower4 = c(2.0508786601119,  1.82531013984407, 2.70909399201808, 2.58177942462658),
                       thetaUpper1 = c(1.20151430505538, 2.5445464439953, 0.789912481680675, 1.51912419362987),
                       thetaUpper2 = c(1.20151430505538, 3.11445464439953, 1.42982495424694, 2.07241846025026),
                       thetaUpper3 = c(1.20151430505538,      4.11445464439953, 2.06973742976804, 2.63297479420376),
                       thetaUpper4 = c(2.08163969688621,         5.06957951511996, 2.72942818864999, 3.21545615944865),
                       wrt = c(NA,        0.0362270956364099, NA, 0.0117852907511687),
                       wint = c(NA, 0.883174518432851,      NA, 0.872504068254756),
                       wx = c(NA, 0.109102515307999, NA, 0.124710987685301    ),
                       b = c(NA, 0.822557962816407, NA, 0.639762190335742),
                       a = c(1.76809729183024,    0.358796422821696, 1.33047749824592, 0.454392270307959),
                       z = c(0.144642348507182,        NA, 0.301234205760964, NA),
                       sz = c(0.117511684791096, NA, 0.120202480392568,   NA),
                       sigvis = c(0.228806568219654, NA, 0.189970554333006, NA),
                       tau = c(1.11663576818929,      NA, 0.989604987185369, NA),
                       w = c(0.296698197166943, NA, 0.489995452399694,    NA),
                       sig = c(0.231412989891031, NA, 0.49023842530622, NA),
                       svis = c(0.0321073472766225,   NA, 0.989920023402767, NA)),
                  row.names = c(NA, -4L), class = "data.frame")

preds <- predictConfModels(fits, subdivisions = 100, simult_conf = FALSE)
test_that("Prediction sums to 1", {
  expect_equal(sum(preds$p), 40, tolerance = 0.005)

})




#as.matrix(jobs)[2,]