RenewGLM-package: Renewable Estimation and Incremental Inference in Generalized...

In luolsph/RenewGLM_pkg: Renewable Estimation and Incremental Inference in Generalized Linear Models with Streaming Datasets

Renewable Estimation and Incremental Inference in Generalized Linear Models with Streaming Datasets

Description

This package updates the regression coefficients and their standard errors in generalized linear models as data batches arrive sequentially.

Details

This package aims to update the regression coefficients as data batches arrive sequentially. There are two main functions in package. RenewGLM.inloop is used to processing a sequence of datasets that are stored in a given directory, and the major input is the name of the directory. RenewGLM.outloop is applied in the case where data batches are imported externally, and the form of the input is X and y of the current data batch.

Author(s)

Lan Luo and Peter X.-K. Song
Maintainer: Lan Luo <[e-mail]>

Examples

#Processing data batches internally
N=1000
B=10
p=5
n=N/B
beta<-c(0.2,-0.2,0.2,-0.2,0.2)

tempdatadir<-"~/Desktop/tempdata"
datagenerator_in(beta=beta,n=n, p=p, B=B, family="binomial", construct="cs",
    rho=0.5, tempdatadir=tempdatadir)
RenewGLM_in(B, tempdatadir=tempdatadir, "binomial", p=p, intercept=TRUE)
unlink(tempdatadir)

#Processing data batches externally
N=1000
B=10
p=5
n=N/B
beta<-c(0.2,-0.2,0.2,-0.2,0.2)
infomats<-diag(0,p,p);
betahat<-rep(0,p)
for(b in 1:10){
 data<-datagenerator_out(beta,b,n,"binomial","cs",0.5)
 y<-data[,1]
 X<-data[,-1]
 summary<-RenewGLM_out(X,y,"binomial",betahat,infomats,intercept=TRUE,s,phi)
 betahat<-summary[[1]]
 infomats<-summary[[2]]
 rm(data)
}
sd<-sqrt(diag(solve(infomats)));
pvalue<-2*pnorm(-abs(betahat)/sd)
result<-cbind(betahat=betahat,sd=sd,pvalue=pvalue)
colnames(result)<-c("Estimates","Std.Errors","p-values")

luolsph/RenewGLM_pkg documentation built on April 17, 2023, 12:36 a.m.