CategorModel: R6 class for fitting and predicting joint probability for a...

In stremr: Streamlined Estimation of Survival for Static, Dynamic and Stochastic Treatment and Monitoring Regimes

Description

This R6 class defines and fits a conditional probability model P(A[j]|W,...) for a univariate categorical summary measure A[j]. This class inherits from GenericModel class. Defines the fitting algorithm for a regression model A[j] ~ W + .... Reconstructs the likelihood P(A[j]=a[j]|W,...) afterwards. Categorical A[j] is first redefined into length(levels) bin indicator variables, where levels is a numeric vector of all unique categories in A[j]. The fitting algorithm estimates the binary regressions for hazard for each bin indicator, Bin_A[j][i] ~ W, i.e., the probability that categorical A[j] falls into bin i, Bin_A[j]_i, given that A[j] does not fall in any prior bins Bin_A[j]_1, ..., Bin_A[j]_{i-1}. The dataset of bin indicators (BinA[j]_1,...,BinA[j]_M) is created inside the passed data or newdata object when defining length(levels) bins for A[j].

Usage

CategorModel

Format

An R6Class generator object

Details

• reg - .

• outvar - .

• levels - .

• nbins - .

Methods

new(reg, DataStorageClass.g0, ...)

...

fit(data)

...

predict(newdata)

...

predictAeqa(newdata)

...

Active Bindings

cats

...

stremr documentation built on May 30, 2017, 6:35 a.m.