MBASICFit-class: An S-4 class containing the model fit information for MBASIC...

In chandlerzuo/mbasic: Matrix Based Analysis for State-space Inference and Clustering.

Description

Theta

A K*S by I matrix. The (k+K*(s-1),i)-th element is the posterior probability for the i-th locus to have state s under the k-th experimental condition.

W

A K*S by J matrix. The (k+K*(s-1),j)-th element is the posterior probability that loci in the j-th group are UNENRICHED under the k-th experimental condition.

V

An N by M matrix. The (n,m)-th element is the weight of the m-th component for the n-th replicate. For each row, entries corresponding to the same state should sum to 1.

Z

An I by J matrix. The (i,j)-th element is the posterior probability that the i-th unit (locus) belongs to the j-th cluster.

clustProb

An I by (J + 1) matrix. The(i, j+1)-th element is the posterior probability that the i-th unit belongs to the j-th cluster, and the (i, 1)-th element is the posterior probability that the i-th unit belongs to the singleton cluster.

b

A vector of length I. The i-th element is the posterior probability that the i-th unit (locus) does not belong to any cluster.

aic

The AIC value of the fitted model.

bic

The BIC value of the fitted model.

aicc

The AICC value of the fitted model.

lik

The log-likelihood after the final iteration.

alllik

The vector for the log-likelihood after each E-M iteration.

zeta

The hyper probability for each unit (locus) to belong to some cluster.

Mu

This is an N by M matrix, where the (n,m)-th entry is the mean parameter for the m-th component for the n-th replicate.

Sigma

An N by M matrix, where the (n,m)-th entry is the dispersion parameter for the m-th component for the n-th replicate.

sigma0

(This slot is deprecated.) NULL except if this object is fitted by the MBASIC.binary function. In that case, this is a vector of length N. The n-th entry is the dispersion parameter for the unenrichment component of the n-th experiment.

e

(This slot is deprecated) NULL except if this object is fitted by the MBASIC.binary function. In that case, this is a vector of length N. The n-th entry is the normalization value for the background of the n-th experiment.

probz

A vector of length J. The j-th entry is the hyper probability of any locus to belong to the j-th cluster.

P

A matrix of length I by S. The (i,s)-th entry is the probability for the i-th locus to have state s, conditional on that this locus does not belong to any cluster.

converged

Whether the final model is converged.

Theta.err

Mean squared error in the Theta matrix.

Mu.err

Mean squared error in the Mu parameter.

ARI

Adjusted Rand Index between the fitted clustering and the true clustering.

W.err

Mean squared error in the W matrix.

MisClassRate

Mis-classification rate by comparing the fitted clustering and the true clustering.

Iter

Number of iterations taken.

Loss

A list object for different terms of loss function values.

Struct

A matrix for the structure constraints for each cluster.

Examples

showClass("MBASICFit")

chandlerzuo/mbasic documentation built on May 13, 2019, 3:24 p.m.