PSP-methods: PSP-methods

In Syksy/ePCR: Ensemble Penalized Cox Regression for Survival Prediction

PSP-methods

Description

PSP-methods

print.PSP: Print general information of PSPs contents to the terminal

plot.PSP: By default the mean CV surface in terms of alpha/lambda is plotted using hamlet-package's hmap-function

coef.PSP: Default PSP coef-function extracts only the optimum parameters, not whole lambda-range

predict.PSP: Predict for a novel patient from current PSP

PSP.KM: Kaplan-Meier with division at a given cutoff point within [0,1]

PSP.PCA: Principal Component Plot of a single PSP, showing 2 principal axes with a colouring if strata have been indicated; newx can also be plotted in relation to fitting data

PSP.BOX: Boxplot of a single variable in a PSP in respect to strata, for outlier detection and observing variable distributions

PSP.CSP: Cumulative survival probabilities

PSP.NA: Nelson-Aalen with time-to-event prediction at point t = F^-1(0.5)

Usage

## S4 method for signature 'PSP'
print(x, ...)

plot(x, y, ...)

## S4 method for signature 'PSP'
plot(x, y, bias = 0.1, ...)

## S4 method for signature 'PSP'
coef(object)

## S4 method for signature 'PSP'
predict(object, type = "response", newx, verb = 0)

PSP.KM(object, ...)

## S4 method for signature 'PSP'
PSP.KM(object, cutoff = 0.5)

PSP.PCA(object, ...)

## S4 method for signature 'PSP'
PSP.PCA(
  object,
  newx,
  expanded = TRUE,
  type = "all",
  shuffle = TRUE,
  z = TRUE,
  cex = 1,
  col = c("aquamarine", "coral", "royalblue", "black"),
  pch = 16
)

PSP.BOX(object, ...)

## S4 method for signature 'PSP'
PSP.BOX(object, newx, var = colnames(object@x)[1], expanded = FALSE)

PSP.CSP(object, ...)

## S4 method for signature 'PSP'
PSP.CSP(object, newx, t = seq(from = 1, to = 36 * 30.5, by = 1), plot = FALSE)

PSP.NA(object, ...)

## S4 method for signature 'PSP'
PSP.NA(object, newx, plot = TRUE)

Arguments

x	Generic x
...	Additional custom parameters passed on
y	Generic y
bias	Bias for skewing the color in heatmap key plotting
object	PSP-object
type	Types of variables to include; recognizes (int)eger, (bin)ary and (num)eric
newx	New data matrix
verb	Level of verbosity
cutoff	Cutoff point for division
expanded	Should data matrix expansion through interactions be included
shuffle	Shuffle plotting order
z	Should data centering and scaling should be conducted
cex	Zooming multiplier
col	Vector of color numbers or names to use for strata
pch	Point type to use (refer to par-function pch-parameter)
var	Name of variable to plot
t	Sequence of time points to evaluate cumulative survival probabilities at
plot	Plot the corresponding functionality

Note

Please refer to the PSP-class examples for applying these PSP-methods

Syksy/ePCR documentation built on Feb. 20, 2024, 10:16 p.m.