DataTemporalMap-class: Class DataTemporalMap

In EHRtemporalVariability: Delineating Temporal Dataset Shifts in Electronic Health Records

Description

Class DataTemporalMap object contains the statistical distributions of data estimated at a specific time period. Both relative and absolute frequencies are included.

Details

Objects of this class are generated automatically by the estimateDataTemporalMap function, but its construction and extension is open towards fostering its use through external methods. E.g., one may use additional probability distribution estimation methods, or even construct compatible DataTemporalMaps for other unstructured data such as images or free text.

Value

A DataTemporalMap object.

Slots

probabilityMap

v-by-d numerical matrix representing the probability distribution temporal map (relative frequency).

countsMap

v-by-d numerical matrix representing the counts temporal map (absolute frequency).

dates

d-dimensional Date array of the temporal batches.

support

v-by-1 numerical or character matrix representing the support (the value at each bin) of probabilityMap and countsMap.

variableName

name of the variable (character).

variableType

type of the variable (character) among "numeric", "character", "Date" and "factor".

period

batching period among "week", "month" and "year".

Examples

# Generation through estimateDataTemporalMap function:
dataset <- read.csv2(system.file("extdata",
                                   "nhdsSubset.csv",
                                   package="EHRtemporalVariability"), 
                     sep  = ",",
                     header = TRUE, 
                     na.strings = "", 
                     colClasses = c( "character", "numeric", "factor",
                                     "numeric" , rep( "factor", 22 ) ) )

datasetFormatted <- EHRtemporalVariability::formatDate(
                     input         = dataset,
                     dateColumn    = "date",
                     dateFormat = "%y/%m")

probMaps <- estimateDataTemporalMap(data = datasetFormatted, 
                     dateColumnName = "date", 
                     period         = "month")

class( probMaps[[1]] ) 

# Manual generation:
countsMatrix <- matrix(sample.int(25, size = 12*10, replace = TRUE), nrow = 12, ncol = 10)
probabilityMatrix <- sweep(countsMatrix,1,rowSums(countsMatrix),"/")
dates <- seq(Sys.Date(),(Sys.Date()+30*11),30)
x <- new('DataTemporalMap', probabilityMap = probabilityMatrix, 
                     countsMap = countsMatrix, dates = dates, support = data.frame(1:10), 
                     variableName = "example", variableType = "numeric", period = "month")
plotDataTemporalMap(x)

EHRtemporalVariability documentation built on May 31, 2021, 5:07 p.m.