This function predicts the discrete choice behaviour.
Usage
# S3 method for class 'RprobitB_fit'
predict(object, data = NULL, overview = TRUE, digits = 2, ...)Arguments
- object
An object of class
RprobitB_fit.- data
Either
NULL, using the data inobject,an object of class
RprobitB_data, for example the test part generated bytrain_test,or a data frame of custom choice characteristics. It must have the same structure as
choice_dataused inprepare_data. Missing columns orNAvalues are set to 0.
- overview
[
logical(1)]
Summarize the prediction in a confusion matrix?- digits
[
integer(1)]
The number of digits of the returned choice probabilities.- ...
Currently not used.
Details
Predictions are made based on the maximum predicted probability for each choice alternative.
See the vignette on choice prediction for a demonstration on how to visualize the model's sensitivity and specificity by means of a receiver operating characteristic (ROC) curve.
Examples
set.seed(1)
data <- simulate_choices(form = choice ~ cov, N = 10, T = 10, J = 2)
data <- train_test(data, test_proportion = 0.5)
model <- fit_model(data$train)
#> Computing sufficient statistics - 0 of 4
#> Computing sufficient statistics - 1 of 4
#> Computing sufficient statistics - 2 of 4
#> Computing sufficient statistics - 3 of 4
#> Computing sufficient statistics - 4 of 4
#> Gibbs sampler - 1 of 1000 iterations
#> Gibbs sampler - 10 of 1000 iterations
#> Gibbs sampler - 20 of 1000 iterations
#> Gibbs sampler - 30 of 1000 iterations
#> Gibbs sampler - 40 of 1000 iterations
#> Gibbs sampler - 50 of 1000 iterations
#> Gibbs sampler - 60 of 1000 iterations
#> Gibbs sampler - 70 of 1000 iterations
#> Gibbs sampler - 80 of 1000 iterations
#> Gibbs sampler - 90 of 1000 iterations
#> Gibbs sampler - 100 of 1000 iterations
#> Gibbs sampler - 110 of 1000 iterations
#> Gibbs sampler - 120 of 1000 iterations
#> Gibbs sampler - 130 of 1000 iterations
#> Gibbs sampler - 140 of 1000 iterations
#> Gibbs sampler - 150 of 1000 iterations
#> Gibbs sampler - 160 of 1000 iterations
#> Gibbs sampler - 170 of 1000 iterations
#> Gibbs sampler - 180 of 1000 iterations
#> Gibbs sampler - 190 of 1000 iterations
#> Gibbs sampler - 200 of 1000 iterations
#> Gibbs sampler - 210 of 1000 iterations
#> Gibbs sampler - 220 of 1000 iterations
#> Gibbs sampler - 230 of 1000 iterations
#> Gibbs sampler - 240 of 1000 iterations
#> Gibbs sampler - 250 of 1000 iterations
#> Gibbs sampler - 260 of 1000 iterations
#> Gibbs sampler - 270 of 1000 iterations
#> Gibbs sampler - 280 of 1000 iterations
#> Gibbs sampler - 290 of 1000 iterations
#> Gibbs sampler - 300 of 1000 iterations
#> Gibbs sampler - 310 of 1000 iterations
#> Gibbs sampler - 320 of 1000 iterations
#> Gibbs sampler - 330 of 1000 iterations
#> Gibbs sampler - 340 of 1000 iterations
#> Gibbs sampler - 350 of 1000 iterations
#> Gibbs sampler - 360 of 1000 iterations
#> Gibbs sampler - 370 of 1000 iterations
#> Gibbs sampler - 380 of 1000 iterations
#> Gibbs sampler - 390 of 1000 iterations
#> Gibbs sampler - 400 of 1000 iterations
#> Gibbs sampler - 410 of 1000 iterations
#> Gibbs sampler - 420 of 1000 iterations
#> Gibbs sampler - 430 of 1000 iterations
#> Gibbs sampler - 440 of 1000 iterations
#> Gibbs sampler - 450 of 1000 iterations
#> Gibbs sampler - 460 of 1000 iterations
#> Gibbs sampler - 470 of 1000 iterations
#> Gibbs sampler - 480 of 1000 iterations
#> Gibbs sampler - 490 of 1000 iterations
#> Gibbs sampler - 500 of 1000 iterations
#> Gibbs sampler - 510 of 1000 iterations
#> Gibbs sampler - 520 of 1000 iterations
#> Gibbs sampler - 530 of 1000 iterations
#> Gibbs sampler - 540 of 1000 iterations
#> Gibbs sampler - 550 of 1000 iterations
#> Gibbs sampler - 560 of 1000 iterations
#> Gibbs sampler - 570 of 1000 iterations
#> Gibbs sampler - 580 of 1000 iterations
#> Gibbs sampler - 590 of 1000 iterations
#> Gibbs sampler - 600 of 1000 iterations
#> Gibbs sampler - 610 of 1000 iterations
#> Gibbs sampler - 620 of 1000 iterations
#> Gibbs sampler - 630 of 1000 iterations
#> Gibbs sampler - 640 of 1000 iterations
#> Gibbs sampler - 650 of 1000 iterations
#> Gibbs sampler - 660 of 1000 iterations
#> Gibbs sampler - 670 of 1000 iterations
#> Gibbs sampler - 680 of 1000 iterations
#> Gibbs sampler - 690 of 1000 iterations
#> Gibbs sampler - 700 of 1000 iterations
#> Gibbs sampler - 710 of 1000 iterations
#> Gibbs sampler - 720 of 1000 iterations
#> Gibbs sampler - 730 of 1000 iterations
#> Gibbs sampler - 740 of 1000 iterations
#> Gibbs sampler - 750 of 1000 iterations
#> Gibbs sampler - 760 of 1000 iterations
#> Gibbs sampler - 770 of 1000 iterations
#> Gibbs sampler - 780 of 1000 iterations
#> Gibbs sampler - 790 of 1000 iterations
#> Gibbs sampler - 800 of 1000 iterations
#> Gibbs sampler - 810 of 1000 iterations
#> Gibbs sampler - 820 of 1000 iterations
#> Gibbs sampler - 830 of 1000 iterations
#> Gibbs sampler - 840 of 1000 iterations
#> Gibbs sampler - 850 of 1000 iterations
#> Gibbs sampler - 860 of 1000 iterations
#> Gibbs sampler - 870 of 1000 iterations
#> Gibbs sampler - 880 of 1000 iterations
#> Gibbs sampler - 890 of 1000 iterations
#> Gibbs sampler - 900 of 1000 iterations
#> Gibbs sampler - 910 of 1000 iterations
#> Gibbs sampler - 920 of 1000 iterations
#> Gibbs sampler - 930 of 1000 iterations
#> Gibbs sampler - 940 of 1000 iterations
#> Gibbs sampler - 950 of 1000 iterations
#> Gibbs sampler - 960 of 1000 iterations
#> Gibbs sampler - 970 of 1000 iterations
#> Gibbs sampler - 980 of 1000 iterations
#> Gibbs sampler - 990 of 1000 iterations
#> Gibbs sampler - 1000 of 1000 iterations
predict(model)
#> predicted
#> true A B
#> A 16 8
#> B 6 20
predict(model, overview = FALSE)
#> id idc A B true predicted correct
#> 1 1 1 0.57 0.43 B A FALSE
#> 2 1 2 0.53 0.47 A A TRUE
#> 3 1 3 0.54 0.46 A A TRUE
#> 4 1 4 0.21 0.79 B B TRUE
#> 5 1 5 0.31 0.69 A B FALSE
#> 6 1 6 0.97 0.03 A A TRUE
#> 7 1 7 0.63 0.37 A A TRUE
#> 8 1 8 0.61 0.39 A A TRUE
#> 9 1 9 0.51 0.49 A A TRUE
#> 10 1 10 0.94 0.06 A A TRUE
#> 11 2 1 0.10 0.90 B B TRUE
#> 12 2 2 0.34 0.66 B B TRUE
#> 13 2 3 0.94 0.06 A A TRUE
#> 14 2 4 0.89 0.11 B A FALSE
#> 15 2 5 0.23 0.77 B B TRUE
#> 16 2 6 0.47 0.53 A B FALSE
#> 17 2 7 0.48 0.52 B B TRUE
#> 18 2 8 0.25 0.75 A B FALSE
#> 19 2 9 0.48 0.52 A B FALSE
#> 20 2 10 0.36 0.64 B B TRUE
#> 21 3 1 0.21 0.79 A B FALSE
#> 22 3 2 0.71 0.29 A A TRUE
#> 23 3 3 0.49 0.51 B B TRUE
#> 24 3 4 0.92 0.08 A A TRUE
#> 25 3 5 0.37 0.63 A B FALSE
#> 26 3 6 0.75 0.25 A A TRUE
#> 27 3 7 0.59 0.41 B A FALSE
#> 28 3 8 0.87 0.13 A A TRUE
#> 29 3 9 0.51 0.49 B A FALSE
#> 30 3 10 0.37 0.63 B B TRUE
#> 31 4 1 0.24 0.76 B B TRUE
#> 32 4 2 0.43 0.57 B B TRUE
#> 33 4 3 0.60 0.40 B A FALSE
#> 34 4 4 0.20 0.80 B B TRUE
#> 35 4 5 0.91 0.09 A A TRUE
#> 36 4 6 0.27 0.73 B B TRUE
#> 37 4 7 0.59 0.41 A A TRUE
#> 38 4 8 0.44 0.56 A B FALSE
#> 39 4 9 0.15 0.85 B B TRUE
#> 40 4 10 0.35 0.65 B B TRUE
#> 41 5 1 0.15 0.85 A B FALSE
#> 42 5 2 0.87 0.13 A A TRUE
#> 43 5 3 0.07 0.93 B B TRUE
#> 44 5 4 0.30 0.70 B B TRUE
#> 45 5 5 0.45 0.55 B B TRUE
#> 46 5 6 0.55 0.45 B A FALSE
#> 47 5 7 0.91 0.09 A A TRUE
#> 48 5 8 0.36 0.64 B B TRUE
#> 49 5 9 0.26 0.74 B B TRUE
#> 50 5 10 0.06 0.94 B B TRUE
predict(model, data = data$test)
#> predicted
#> true A B
#> A 23 7
#> B 7 13
predict(
model,
data = data.frame("cov_A" = c(1, 1, NA, NA), "cov_B" = c(1, NA, 1, NA)),
overview = FALSE
)
#> Checking for missing covariates
#> id idc A B prediction
#> 1 1 1 0.56 0.44 A
#> 2 2 1 0.30 0.70 B
#> 3 3 1 0.80 0.20 A
#> 4 4 1 0.56 0.44 A