destiny-deprecated: Compatibility versions of functions for destiny 1.x scripts

In destiny: Creates diffusion maps

Description

This is most useful if you have old scripts that use the old argument order and the default being global sigma. Simply put DiffusionMap <- DiffusionMapCompat on top of your script to make it work with destiny 2.x.

Usage

DiffusionMapCompat(
  data,
  sigma = "global",
  k = find_dm_k(nrow(data) - 1L),
  n.eigs = min(20L, nrow(data) - 2L),
  density.norm = TRUE,
  ...,
  distance = c("euclidean", "cosine", "rankcor"),
  censor.val = NULL,
  censor.range = NULL,
  missing.range = NULL,
  vars = NULL,
  verbose = !is.null(censor.range)
)

cube.helix(
  n = 6,
  start = 0,
  r = 0.4,
  hue = 0.8,
  gamma = 1,
  light = 0.85,
  dark = 0.15,
  reverse = FALSE
)

dm.predict(dm, new.data, verbose = FALSE)

eig.decomp(M, n.eigs, sym = isSymmetric(M))

find.dm.k(n, min.k = 100L, small = 1000L, big = 10000L)

find.sigmas(
  data,
  step.size = 0.1,
  steps = 10L,
  start = NULL,
  sample.rows = 500L,
  early.exit = FALSE,
  ...,
  censor.val = NULL,
  censor.range = NULL,
  missing.range = NULL,
  vars = NULL,
  verbose = TRUE
)

optimal.sigma(object)

lWhich(idx, nms = seq_len(len), len = length(nms), useNames = TRUE)

plot_gradient_map(
  coords,
  exprs,
  ...,
  gene,
  dims = 1:2,
  pal = hcl.colors,
  faceter = facet_wrap(~Gene)
)

Arguments

data, n, n.eigs, start, censor.val, censor.range, missing.range, vars, verbose, ...	Corresponding parameters in multiple functions
sigma, k, density.norm, distance	Corresponding DiffusionMap parameters
r, hue, gamma, light, dark, reverse	Corresponding cube_helix parameters
dm, new.data	Corresponding dm_predict parameters
M, sym	Corresponding eig_decomp parameters
min.k, small, big	Corresponding find_dm_k parameters
step.size, steps, sample.rows, early.exit	Corresponding find_sigmas parameters
object	Corresponding optimal_sigma parameters

Examples

DiffusionMap <- DiffusionMapCompat
## Not run: source('my-old-script.r')

destiny documentation built on Nov. 8, 2020, 7:38 p.m.