fix another mistake in find and replace

R/estimate_mnhmm.R

@@ -33,8 +33,10 @@
 #'                    states = mvad_scodes, labels = mvad_labels, xtstep = 6,
 #'                    cpal = unname(colorpalette[[6]]))
 #' 
+#' \dontrun{
 #' set.seed(1)
 #' fit <- estimate_mnhmm(mvad_seq, n_states = 3, n_clusters = 2)
+#' }
 estimate_mnhmm <- function(
     observations, n_states, n_clusters, initial_formula = ~1, 
     transition_formula = ~1, emission_formula = ~1, cluster_formula = ~1,

R/estimate_nhmm.R

@@ -67,7 +67,9 @@
 #'                    cpal = unname(colorpalette[[6]]))
 #' 
 #' set.seed(1)
+#' \dontrun{
 #' fit <- estimate_nhmm(mvad_seq, n_states = 3)
+#' }
 estimate_nhmm <- function(
     observations, n_states, initial_formula = ~1, 
     transition_formula = ~1, emission_formula = ~1, 

R/forwardBackward.R

@@ -249,7 +249,7 @@ forward_backward.mnhmm <- function(model, forward_only = FALSE,
     model$n_clusters,
     model$n_channels > 1
   )
-  theta_raw <- model$coefficientstheta_raw
+  theta_raw <- model$coefficients$theta_raw
   out <- list()
   if (model$n_channels == 1) {
     out$forward_probs <- forward_mnhmm_singlechannel(

R/get_probs.R

@@ -118,7 +118,7 @@ get_probs.mnhmm <- function(model, newdata = NULL, nsim = 0,
   X_transition <- aperm(model$X_transition, c(3, 1, 2))
   X_emission <- aperm(model$X_emission, c(3, 1, 2))
   X_cluster <- t(model$X_cluster)
-  theta_raw <- model$coefficientstheta_raw
+  theta_raw <- model$coefficients$theta_raw
   initial_probs <- vector("list", D)
   transition_probs <- vector("list", D)
   emission_probs <- vector("list", D)

R/hidden_paths.R

@@ -126,7 +126,7 @@ hidden_paths.mnhmm <- function(model, respect_void = TRUE, ...) {
     model$n_clusters,
     model$n_channels > 1
   )
-  theta_raw <- model$coefficientstheta_raw
+  theta_raw <- model$coefficients$theta_raw
   if (model$n_channels == 1) {
     out <- viterbi_mnhmm_singlechannel(
       beta_i_raw, X_initial,

R/sample_parameters.R

@@ -36,7 +36,7 @@ sample_parameters <- function(model, nsim, probs, return_samples = FALSE) {
   p_s <- length(beta_s_raw)
   p_o <- length(beta_o_raw)
   if (mixture) {
-    theta_raw <- model$coefficientstheta_raw
+    theta_raw <- model$coefficients$theta_raw
     pars <- c(pars, theta_raw)
     if (mixture) p_d <- length(theta_raw)
   }

tests/testthat/test-build_mnhmm.R

@@ -38,7 +38,7 @@ test_that("estimate_mnhmm returns object of class 'mnhmm'", {
       "y", s, d, initial_formula = ~ x, transition_formula = ~z,
       emission_formula = ~ z, cluster_formula = ~ x,
       data = data, time = "time", id = "id", 
-      iter = 0),
+      iter = 0, verbose = FALSE, hessian = FALSE),
     NA
   )
   expect_s3_class(

tests/testthat/test-build_nhmm.R

@@ -22,7 +22,8 @@ test_that("build_nhmm returns object of class 'nhmm'", {
     model <- build_nhmm(
       obs, s, initial_formula = ~ x, transition_formula = ~z,
       emission_formula = ~ z, data = data, 
-      time = "time", id = "id", state_names = 1:s, channel_names = "obs"
+      time = "time", id = "id", state_names = 1:s, channel_names = "obs", 
+      verbose = FALSE
     ),
     NA
   )
@@ -33,10 +34,10 @@ test_that("build_nhmm returns object of class 'nhmm'", {
 })
 test_that("estimate_nhmm returns object of class 'nhmm'", {
   expect_error(
-    capture.output(fit <- estimate_nhmm(
+    fit <- estimate_nhmm(
       "y", s, initial_formula = ~ x, transition_formula = ~z,
       emission_formula = ~ z, data = data, time = "time", id = "id", 
-      iter = 0)),
+      iter = 0, verbose = FALSE, hessian = FALSE),
     NA
   )
   expect_s3_class(

tests/testthat/test-forward_backward.R

@@ -40,7 +40,7 @@ test_that("'forward_backward' works for multichannel 'nhmm'", {
       inits = hmm_biofam[
         c("initial_probs", "transition_probs", "emission_probs")
       ],
-      iter = 1, verbose = FALSE, hessian = FALSE,
+      iter = 1, verbose = FALSE, hessian = FALSE
     ),
     NA
   )
@@ -93,7 +93,7 @@ test_that("'forward_backward' works for multichannel 'mnhmm'", {
   expect_error(
     fit <- estimate_mnhmm(
       hmm_biofam$observations, n_states = 3, n_clusters = 2,
-      iter = 1, verbose = FALSE, hessian = FALSE,
+      iter = 1, verbose = FALSE, hessian = FALSE
     ),
     NA
   )

tests/testthat/test-hidden_paths.R

@@ -26,7 +26,7 @@ test_that("'hidden_paths' works for 'nhmm'", {
       inits = hmm_biofam[
         c("initial_probs", "transition_probs", "emission_probs")
       ],
-      iter = 1, verbose = FALSE, hessian = FALSE,
+      iter = 1, verbose = FALSE, hessian = FALSE
     ),
     NA
   )
@@ -59,7 +59,7 @@ test_that("'hidden_paths' works for 'mnhmm'", {
   expect_error(
     fit <- estimate_mnhmm(
       hmm_biofam$observations, n_states = 3, n_clusters = 2,
-      iter = 1, verbose = FALSE, hessian = FALSE,
+      iter = 1, verbose = FALSE, hessian = FALSE
     ),
     NA
   )

tests/testthat/test-posterior_probs.R

@@ -42,7 +42,7 @@ test_that("'posterior_probs' works for 'nhmm'", {
       inits = hmm_biofam[
         c("initial_probs", "transition_probs", "emission_probs")
       ],
-      iter = 1, verbose = FALSE, hessian = FALSE,
+      iter = 1, verbose = FALSE, hessian = FALSE
     ),
     NA
   )
@@ -74,7 +74,7 @@ test_that("'posterior_probs' works for 'mnhmm'", {
   expect_error(
     fit <- estimate_mnhmm(
       hmm_biofam$observations, n_states = 3, n_clusters = 2,
-      iter = 1, verbose = FALSE, hessian = FALSE,
+      iter = 1, verbose = FALSE, hessian = FALSE
     ),
     NA
   )

tests/testthat/test-simulate_mnhmm.R

@@ -1,4 +1,4 @@
-test_that("simulate_mnhmm works", {
+test_that("simulate_mnhmm, coef and get_probs works", {
   set.seed(1)
   p <- 50
   n <- 10
@@ -35,5 +35,24 @@ test_that("simulate_mnhmm works", {
           "*", "%")), names = ""), class = "table")
 
   )
+  expect_error(
+    fit <- estimate_mnhmm(
+      sim$observations, n_states = 2, 
+      n_clusters = 3, 
+      initial_formula = ~1, transition_formula = ~ x, 
+      emission_formula = ~ x + z, cluster_formula = ~w, 
+      data = d, time = "month", id = "person", 
+      init = sim$model$coefficients, 
+      iter = 1, verbose = FALSE, hessian = FALSE),
+    NA
+  )
+  expect_error(
+    cf <- coef(fit),
+    NA
+  )
+  expect_error(
+    p <- get_probs(fit, nsim = 0),
+    NA
+  )
 })