Rtreemix (version 1.34.0)

RtreemixStats-class: Class "RtreemixStats"

Description

The RtreemixStats class contains the (weighted, log) likelihoods for a given dataset (specified by the parent class) derived from the probability distribution induced by an underlying mutagenetic trees mixture model.

Objects from the Class

Objects can be created by calls of the form new("RtreemixStats", Data, Model, LogLikelihoods, WLikelihoods). The class RtreemixStats extends the RtreemixData class and specifies (log, weighted) likelihoods for these data derived from a given RtreemixModel. The number of the genetic events in the patterns from the given dataset (Data) has to be equal to the number of genetic events in the branchings from the mixture model given by the slot Model. When having the weighted likelihoods, one can easily derive the responsibilities of the model components in Model for generating the patterns in the specified dataset (Data). The Data is an RtreemixData object that specifies the patterns for which the likelihoods are calculated. The Model is an RtreemixModel object that specifies the mutagenetic trees mixture model used for deriving the likelihoods of the given data. The LogLikelihoods is a numeric vector that contains the log-likelihoods of the patterns in Data. Its length equals the sample size, i.e. the number of patients in Data. The WLikelihoods is a numeric matrix that specifies the weighted likelihoods of each pattern in the given dataset Data. The number of columns in WLikelihoods equals the number of tree components in Model and the number of rows equals the number of patients in Data.

Slots

Model:

Object of class "RtreemixModel".

LogLikelihoods:

Object of class "numeric". The length of LogLikelihoods must be equal to the number of patients of the dataset specified with the parent "RtreemixData" class.

WLikelihoods:

Object of class "matrix". The number of rows must be equal to the sample size of the dataset specified with the parent "RtreemixData" class. The number of columns must be identical with the number of tree components in the mixture model Model.

Extends

Class "RtreemixData", directly.

Methods

LogLikelihoods

signature(object = "RtreemixStats"): A method for obtaining the log-likelihoods of the patterns in the dataset specified with the parent "RtreemixData" class.

Model

signature(object = "RtreemixStats"): A method for obtaining the mutagenetic trees mixture model used for deriving the likelihoods.

WLikelihoods

signature(object = "RtreemixStats"): A method for obtaining the weighted likelihoods of the patterns in the dataset specified with the parent "RtreemixData" class.

getData

signature(object = "RtreemixStats"): A method for obtaining the dataset specified with the parent "RtreemixData" class.

getResp

signature(object = "RtreemixStats"): A method for computing the matrix of responsibilities for the trees to generate each of the samples in the parent dataset from their weighted likelihoods WLikelihoods.

References

Learning multiple evolutionary pathways from cross-sectional data, N. Beerenwinkel et al.

See Also

RtreemixData-class, RtreemixModel-class, fit-methods, likelihoods-methods

Examples

## Generate a random RtreemixModel object with 3 components and 9 genetic events.
mod <- generate(K = 3, no.events = 9, noise.tree = TRUE, prob = c(0.2, 0.8))
show(mod)

## Draw a data sample from the model mod.
data <- sim(model = mod, no.draws = 400)

## Create an RtreemixStats object.
mod.stat <- likelihoods(model = mod, data = data)
show(mod.stat)

## See the slots from the RtreemixStats object.
Model(mod.stat)
LogLikelihoods(mod.stat)
WLikelihoods(mod.stat)
## See data.
getData(mod.stat)
## Calculate the responsibilities from the weighted likelihoods.
getResp(mod.stat)