library(pRolocdata)
data("tan2009r1")
set.seed(1)
tansim <- sim_dynamic(object = tan2009r1,
numRep = 6L,
numDyn = 100L)
gpParams <- lapply(tansim$lopitrep, function(x)
fitGPmaternPC(x, hyppar = matrix(c(0.5, 1, 100), nrow = 1)))
d1 <- tansim$lopitrep
control1 <- d1[1:3]
treatment1 <- d1[4:6]
test_that("differential localisation computation", {
set.seed(1)
mcmc1 <- bandle(objectCond1 = control1,
objectCond2 = treatment1,
gpParams = gpParams,
fcol = "markers",
numIter = 10L, burnin = 1L, thin = 2L, numChains = 2,
BPPARAM = SerialParam(RNGseed = 1))
mcmc1 <- bandleProcess(mcmc1)
dp1 <- diffLocalisationProb(mcmc1)
expect_length(dp1, length(rownames(unknownMSnSet(object = tan2009r1, fcol = "markers"))))
expect_true(all(dp1 <= 1))
expect_true(all(dp1 >= 0))
})
test_that("differential localisation computation 2", {
set.seed(1)
mcmc1 <- bandle(objectCond1 = control1,
objectCond2 = treatment1,
gpParams = gpParams,
fcol = "markers",
numIter = 10L, burnin = 1L, thin = 2L, numChains = 2,
BPPARAM = SerialParam(RNGseed = 1))
mcmc1 <- bandleProcess(mcmc1)
.top <- 20
.bootsample <- 100
bdp <- bootstrapdiffLocprob(mcmc1, top = .top, Bootsample = .bootsample,
decreasing = TRUE)
expect_length(bdp[1,], .bootsample)
expect_length(bdp[,1], .top)
expect_equal(order(bdp[,1], decreasing = TRUE), seq.int(.top))
})
test_that("differential localisation computation 3", {
set.seed(1)
mcmc1 <- bandle(objectCond1 = control1,
objectCond2 = treatment1,
gpParams = gpParams,
fcol = "markers",
numIter = 10L, burnin = 1L, thin = 2L, numChains = 2,
BPPARAM = SerialParam(RNGseed = 1))
mcmc1 <- bandleProcess(mcmc1)
.top <- 20
.nsample <- 100
dp <- binomialDiffLocProb(mcmc1, top = .top, nsample = .nsample,
decreasing = TRUE)
probs <- diffLocalisationProb(params = mcmc1)
probs <- probs[order(probs, decreasing = TRUE)]
expect_length(dp[1,], .nsample)
expect_length(dp[,1], .top)
expect_equal(names(probs)[seq.int(.top)], rownames(dp))
})
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.