inst/tinytest/test_synthetic_data.R

# test manifold_synthetic_data -------------------------------------------------

n_samples <- 1000L

## Swiss roll ------------------------------------------------------------------

swiss <- manifold_synthetic_data(
  type = "swiss_roll",
  n_samples = n_samples,
  seed = 42L
)

expect_true(
  current = checkmate::testMatrix(
    swiss$data,
    mode = "numeric",
    nrow = n_samples
  ),
  info = "swiss_roll returns numeric matrix with correct number of rows"
)

expect_null(
  current = swiss$membership,
  info = "swiss_roll membership is NULL"
)

## clusters --------------------------------------------------------------------

n_clusters <- 5L

clusters <- manifold_synthetic_data(
  type = "clusters",
  n_samples = n_samples,
  dim = 10L,
  seed = 42L,
  parameters = params_clusters(n_clusters = n_clusters)
)

expect_true(
  current = checkmate::testMatrix(
    clusters$data,
    mode = "numeric",
    nrow = n_samples,
    ncol = 10L
  ),
  info = "clusters returns numeric matrix with correct dimensions"
)

expect_true(
  current = checkmate::testInteger(clusters$membership, len = n_samples),
  info = "clusters membership is integer vector of correct length"
)

expect_equal(
  current = length(unique(clusters$membership)),
  target = n_clusters,
  info = "clusters membership has correct number of unique clusters"
)

## trajectory ------------------------------------------------------------------

### pre-defined topologies -----------------------------------------------------

topologies <- c("bifurcation", "linear", "combination")

for (topo in topologies) {
  traj <- manifold_synthetic_data(
    type = "trajectory",
    n_samples = n_samples,
    dim = 10L,
    seed = 42L,
    parameters = params_trajectory(topology = topo)
  )

  expect_true(
    current = checkmate::testMatrix(
      traj$data,
      mode = "numeric",
      nrow = n_samples,
      ncol = 10L
    ),
    info = sprintf(
      "%s topology returns numeric matrix with correct dimensions",
      topo
    )
  )

  expect_true(
    current = checkmate::testInteger(traj$membership, len = n_samples),
    info = sprintf(
      "%s topology membership is integer vector of correct length",
      topo
    )
  )
}

### custom cell_trajectories ---------------------------------------------------

custom_trajectories <- list(
  parent = c(NA_integer_, 0L, 0L),
  split_at = c(0.0, 0.5, 0.5),
  length = c(1.0, 1.0, 1.0)
)

traj_custom <- manifold_synthetic_data(
  type = "trajectory",
  n_samples = n_samples,
  dim = 10L,
  seed = 42L,
  parameters = params_trajectory(cell_trajectories = custom_trajectories)
)

expect_true(
  current = checkmate::testMatrix(
    traj_custom$data,
    mode = "numeric",
    nrow = n_samples,
    ncol = 10L
  ),
  info = "custom cell_trajectories returns numeric matrix with correct dimensions"
)

expect_true(
  current = checkmate::testInteger(traj_custom$membership, len = n_samples),
  info = "custom cell_trajectories membership is integer vector of correct length"
)

## hierarchical ----------------------------------------------------------------

n_supergroups <- 3L
n_subclusts <- 4L

hier <- manifold_synthetic_data(
  type = "hierarchical",
  n_samples = n_samples,
  dim = 10L,
  seed = 42L,
  parameters = params_hierarchical(
    n_supergroups = n_supergroups,
    n_subclusts = n_subclusts
  )
)

expect_true(
  current = checkmate::testMatrix(hier$data, mode = "numeric", ncol = 10L),
  info = "hierarchical returns numeric matrix with correct number of columns"
)

expect_true(
  current = checkmate::testInteger(hier$membership$supergroup),
  info = "hierarchical supergroup membership is an integer vector"
)

expect_true(
  current = checkmate::testInteger(hier$membership$subgroup),
  info = "hierarchical subgroup membership is an integer vector"
)

expect_equal(
  current = length(unique(hier$membership$supergroup)),
  target = n_supergroups,
  info = "hierarchical membership has correct number of unique supergroups"
)

expect_equal(
  current = length(unique(hier$membership$subgroup)),
  target = n_supergroups * n_subclusts,
  info = "hierarchical membership has correct number of unique subgroups"
)

expect_equal(
  current = length(hier$membership$supergroup),
  target = length(hier$membership$subgroup),
  info = "hierarchical supergroup and subgroup membership vectors are the same length"
)

## input validation ------------------------------------------------------------

expect_error(
  current = manifold_synthetic_data(type = "swiss_roll", n_samples = -1L),
  info = "negative n_samples raises error"
)

expect_error(
  current = manifold_synthetic_data(
    type = "clusters",
    n_samples = n_samples,
    dim = 0L
  ),
  info = "dim of 0 raises error"
)

expect_error(
  current = manifold_synthetic_data(
    type = "clusters",
    n_samples = n_samples,
    parameters = params_hierarchical()
  ),
  info = "wrong params type for chosen data type raises error"
)

expect_error(
  current = manifold_synthetic_data(
    type = "trajectory",
    n_samples = n_samples,
    parameters = list(topology = "bifurcation", noise = 0.1)
  ),
  info = "plain list missing required fields raises error"
)

expect_error(
  current = manifold_synthetic_data(
    type = "trajectory",
    n_samples = n_samples,
    parameters = params_trajectory(
      cell_trajectories = list(parent = NA_integer_, split_at = 0.5)
    )
  ),
  info = "malformed cell_trajectories missing 'length' raises error"
)

expect_error(
  current = manifold_synthetic_data(
    type = "trajectory",
    n_samples = n_samples,
    parameters = params_trajectory(
      cell_trajectories = list(
        parent = c(NA_integer_, 0L),
        split_at = c(0.0),
        length = c(1.0, 1.0)
      )
    )
  ),
  info = "cell_trajectories with unequal vector lengths raises error"
)

Try the manifoldsR package in your browser

Any scripts or data that you put into this service are public.

manifoldsR documentation built on Nov. 17, 2026, 5:09 p.m.