modelFitStatistics: Print statistics of a c3co model

In pneuvial/c3co: Cancer Cell Clonality using Copy-Number

Description

Print statistics of a c3co model

Usage

modelFitStatistics(Y, Yhat, What, Zhatt)

Arguments

Y	Original n-by-J matrix of data for n samples and J segments.
Yhat	Estimated n-by-J matrix for n samples and J segments.
What	Inferred n-by-K weight matrix for n samples and K subclones.
Zhatt	The transposed version of the inferred n-by-J subclones matrix 'Zhat' for n samples and J segments.

Value

A named numeric vector with the main statistics of the inferred model:

• BIC : Bayesian Information Criterion

• PVE : Percentage of Variation Explained (Nowak et al., 2011)

• logLik: Log Likelihood

• loss : Loss, which equals RSS / (n * J), where RSS is the residual sum of squares

References

Nowak, G., Hastie, T., Pollack, J. R., & Tibshirani, R. (2011). A fused lasso latent feature model for analyzing multi-sample aCGH data. Biostatistics, 12(4), 776-791

pneuvial/c3co documentation built on May 25, 2019, 10:21 a.m.