BinaryOutcomeModel: R6 class for fitting and making predictions for a single binary outcome regression model P(B | PredVars)
Description
This R6 class can request, store and manage the design matrix Xmat, as well as the binary outcome Bin for the
logistic regression P(Bin|Xmat).
Can also be used for converting data in wide format to long when requested,
e.g., when pooling across binary indicators (fitting one pooled logistic regression model for several indicators)
The class has methods that perform queries to data storage R6 class DataStorageClass to get appropriate data columns & row subsets
Details
- cont.sVar.flag - Is the original outcome variable continuous?
- bw.j - Bin width (interval length) for an outcome that is a bin indicator of a discretized continous outcome.
- GLMpackage - Controls which package will be used for performing model fits (
glm or speedglm).
- binomialModelObj - Pointer to an instance of
binomialModelObj class that contains the data.