Description Usage Arguments Details Value Examples
EM algorithm to estimate the copula mixture model for discrete data
1 2 | est.IDR.discrete(x, y, mu, sigma, rho, p, eps, n.missing, miss.sym,
as.single.loglik, as.single.em, common.only = TRUE, labels = NULL)
|
x |
replicate 1 : a vector of values from original observation |
y |
replicate 2 : paired with |
mu |
a starting value for the mean of the reproducible component. |
sigma |
a starting value for the standard deviation of the reproducible component. |
rho |
a starting value for the correlation coefficient of the reproducible component. |
p |
a starting value for the proportion of reproducible component. |
eps |
a small number to control convergence. For example, |
n.missing |
|
miss.sym |
symbol for missing value. For example, |
as.single.loglik |
an integer n to control the precision of numberical integration. If n=1, the computation is exact. If n>1, only integrate the bins with counts more than n, and treat bins with counts <= n as singletons in the likelihood computation |
as.single.em |
similar to |
common.only |
If |
labels |
the cdf on the left side of the lower boundary of the bins |
EM to compute the latent structure steps: 1. raw values are first transformed into pseudovalues 2. EM is used to compute the underlining structure, which is a mixture of two normals
para estimated parameters: p, rho, mu, sigma, estimated n.missing.
loglik can be used for mapping category format back to original
e.g. x.cat[level.factor[1]] == x[1], y.cat[level.factor[1]]==y[1]
.
loglik.trace trajectory of log-likelihood.
idr.cat a numeric vector of the local idr for each category (i.e. estimated conditional probablility for each observation to belong to the irreproducible component).
IDR.cat a numerical vector of the expected irreproducible discovery rate for categories that are as irreproducible or more irreproducible than the given categories.
idr.obs a numeric vector of the local idr for each observation.
IDR.obs a numerical vector of the expected irreproducible discovery rate for observations that are as irreproducible or more irreproducible than the given observations.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 | #load chip_seq data
data(chip_seq)
x = chip_seq[,1]
y = chip_seq[,2]
# Initiation
mu <- 2.6
sigma <- 1.3
rho <- 0.8
p <- 0.7
eps <- 0.001
n.missing <- 0
# Estimate parameters of mixture model
gidr.out <- est.IDR.discrete(x, y, mu, sigma, rho, p, eps, n.missing,
miss.sym = 0, as.single.loglik = 1,
as.single.em = 1, common.only=TRUE, labels=NULL)
names(gidr.out)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.