estimateBP: Estimate parameters of a beta-Poisson model for a data vector

In nghiavtr/BPSC: Beta-Poisson Model for Single-Cell RNA-seq Data Analyses

Description

Estimate parameters of a beta-Poisson model for a data vector

Usage

estimateBP(x, para.num = 4, tbreak.num = 10, break.thres = 10,
  useExt = FALSE, param0 = NULL)

Arguments

x	A vector of input data points
para.num	Mode of beta-Poisson model: 3, 4 (default) or 5 parameters
tbreak.num	Number of breaks for binning
break.thres	A parameter setting of getTbreak function
useExt	A parameter setting of getTbreak function that allows to extend the last bin to infinity or not
param0	Initial parameters for the model. If it is not set, default initial parameters will be obtained from getInitParam function

Value

An optimal model to the input data including optimal parameters (par), X2 test results (X2 and PVAL), etc..

Examples

set.seed(2015)
#Create a simulated gene expression by randomly generating 100 data points
#from a beta-poisson model
alp=0.6;bet=1.5;lam1=20;lam2=0.05
par0=c(alp,bet,lam1,lam2)
bp.vec=rBP(100,par0)
#Estimate parameters of the four-parameter beta-Poisson model from the data points
res=estimateBP(bp.vec,para.num=4)
#Print the goodness-of-fit of the model and the optimal parameters
res$X2
res$par

nghiavtr/BPSC documentation built on May 23, 2019, 4:42 p.m.