Calculate the log CATE score given the baseline covariates and follow-up time for specified scoring method methods

Based on intxcount results of the CATE coefficients estimated with boosting, naive Poisson, two regression, contrast regression, negative binomial

Usage

scorecount(
  fit,
  x.cate,
  time,
  score.method = c("boosting", "poisson", "twoReg", "contrastReg", "negBin")
)

Arguments

fit

List of objects generated from intxcount: outputs of boosting, naive Poisson, two regression, contrast regression, negative binomial

x.cate

Matrix of p.cate baseline covariates; dimension n (observations) by p.cate (covariates in the outcome model)

time

Log-transformed person-years of follow-up; vector of size n

score.method

A vector of one or multiple methods to estimate the CATE score. Allowed values are: 'boosting', 'poisson', 'twoReg', 'contrastReg', 'negBin'. Default specifies all 5 methods.

Value

score.boosting: Estimated log CATE score for all n observations with the boosting method; vector of size n score.poisson: Estimated log CATE score for all n observations with the naive Poisson method; vector of size n score.twoReg: Estimated log CATE score for all n observations with the two regression method; vector of size n score.contrastReg: Estimated log CATE score for all n observations with the contrast regression method; vector of size n score.negBin: Estimated log CATE score for all n observations with the naive Poisson method; vector of size n score = NA if the corresponding method is not called