DataStorageClass: R6 class for storing, managing, subsetting and manipulating...

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

Description

The class DataStorageClass is the only way the package uses to access the input data. The evaluated summary measures from sVar.object are stored as a matrix (private$.mat.sVar). Contains methods for replacing missing values with default in gvars$misXreplace. Also contains method for detecting / setting sVar variable type (binary, categor, contin). Contains methods for combining, subsetting, discretizing & binirizing summary measures (sW,sA). For continous sVar this class provides methods for detecting / setting bin intervals, normalization, disretization and construction of bin indicators. The pointers to this class get passed on to GenericModel functions: $fit(), $predict() and $predictAeqa().

Usage

DataStorageClass

Format

An R6Class generator object

Details

• YnodeVals

• det.Y

Methods

new(Odata, nodes, YnodeVals, det.Y, ...)

...

def.types.sVar(type.sVar = NULL)

...

fixmiss_sVar()

...

set.sVar(name.sVar, new.type)

...

set.sVar.type(name.sVar, new.type)

...

get.sVar(name.sVar, new.sVarVal)

...

replaceOneAnode(AnodeName, newAnodeVal)

...

replaceManyAnodes(Anodes, newAnodesMat)

...

addYnode(YnodeVals, det.Y)

...

evalsubst(subset_exprs, subset_vars)

...

get.dat.sVar(rowsubset = TRUE, covars)

...

get.outvar(rowsubset = TRUE, var)

...

bin.nms.sVar(name.sVar, nbins)

...

detect.sVar.intrvls(name.sVar, nbins, bin_bymass, bin_bydhist, max_nperbin)

...

detect.cat.sVar.levels(name.sVar)

...

binirize.sVar(name.sVar, ...)

...

get.sVar.bw(name.sVar, intervals)

...

get.sVar.bwdiff(name.sVar, intervals)

...

Active Bindings

nobs

...

ncols.sVar

...

names.sVar

...

type.sVar

Named list of length ncol(private$.mat.sVar) with sVar variable types: "binary"/"categor"/"contin".

dat.sVar

...

ord.sVar

Ordinal (categorical) transformation of a continous covariate sVar.

active.bin.sVar

Name of active binarized cont sVar, changes as fit/predict is called (bin indicators are temp. stored in private$.mat.bin.sVar)

dat.bin.sVar

...

emptydat.sVar

...

emptydat.bin.sVar

...

noNA.Ynodevals

...

nodes

...

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