calCurvPlot {ClinicalUtilityRecal}

Calibration Curve Plot

Description

Flexible function for plotting the calibration curve for a risk score and different recalibrations of a risk score. Histograms of risk score and recalibrated risk scores can be plotted using argument "hist==TRUE". Dotted guidelines for the risk threshold R are plotted and clinically releavant risk region [R_l, R_u] are plotted to help gauge calibration in clinically relevants.

Usage

calCurvPlot(y,p,p.std=NULL,p.recal=NULL,
            stdPlot=FALSE, recalPlot=FALSE,
            xlim=c(0,1),ylim=c(0,1),
            label="Original Risk Score",
            label2 = "Standard Recalibrated Risk Score",
            label3 = "Weighted/Constrained Recalibrated Risk Score",
            legendLab = c("Orig.", "Std.", "Wt."),
            mainTitle="Calibration of Risk Score",
            hist=TRUE,ylimHist = c(0,0.5),
            r,rl = -Inf, ru = Inf)

Arguments

y

Vector of binary outcomes, with 1 indicating event (case) and 0 indicating no event (controls)

p

Vector of risk score values

p.std

Vector of risk score values after standard logistic recalibration

p.recal

Vector of risk score values after weighted/contratined logistic recalibration

stdPlot

If TRUE plot calibration curve for standard logistic recalibrated risk score

recalPlot

If TRUE plot calibration curve for weighted/contratined recalibrated risk score

xlim

Limits for x-axis

ylim

Limits for y-axis

label

Label for x-axis corresponding to p vector

label2

Label for x-axis corresponding to p.std vector

label3

Label for x-axis corresponding to p.recal vector

legendLab

Label for legend

mainTitle

Main title for plot

hist

If true plot distribution of risk scores along with calibration curve

ylimHist

Limits for y-axis of histogram

r

Clinically relevant risk threshold used indicated with dotted line

rl

Lower bound of clinically relevant region indicated with dotted-dashed line

ru

Upper bound of clinically relevant region indicated with dotted-dashed line

Value

Prints calibration plot of risk model p, and optional histogram of risk score

Author(s)

Anu Mishra

References

Mishra, A. (2019). Methods for Risk Markers that Incorporate Clinical Utility (Doctoral dissertation). (Available Upon Request)

Examples

## Not run: 
data("fakeData")

### get standard recalibrated risk score
stdRecal.res <- stdRecal(y = fakeData$y,p = fakeData$p)
p.std <- stdRecal.res$p.std

### Select tuning parameter lamba using 5-fold cross-validation repeated 25 times
grid <- RAWgrid(r = 0.3,rl = -Inf,ru = Inf,p = fakeData$p,y = fakeData$y,
                cvParm = "lambda",rl.raw = 0.25,ru.raw = 0.35)
repCV <- cvRepWtTuning(y = fakeData$y,p = fakeData$p,rl = -Inf,ru = Inf,r = 0.3,
                       kFold = 5,cvRep = 25,cvParm = "lambda",tuneSeq = grid,stdErrRule = TRUE)

## Implement weighted logistic recalibration
wtRecal.res <- wtRecal(y = fakeData$y,p = fakeData$p,r = 0.3,rl = -Inf,ru = Inf,
                       lambda = repCV$cv.lambda,delta=1)
p.recal <- wtRecal.res$p.wt

### Calibration curve of only original risk score with histogram
calCurvPlot(y=fakeData$y,p=fakeData$p,p.std=NULL,p.recal=NULL,
            stdPlot=FALSE, recalPlot=FALSE,
            xlim=c(0,1),ylim=c(0,1),
            label="Original Risk Score",
            label2 = "Standard Recalibrated Risk Score",
            label3 = "Weighted/Constrained Recalibrated Risk Score",
            legendLab = c("Orig.", "Std.", "Wt."),
            mainTitle="Calibration of Risk Score",
            hist=TRUE,ylimHist = c(0,0.5),
            r=0.3,rl = -Inf, ru = Inf)

### Calibration curve of only original, standard and weighted recalibrated risk score
calCurvPlot(fakeData$y,p=fakeData$p,p.std=p.std,p.recal=p.recal,
            stdPlot=TRUE, recalPlot=TRUE,
            xlim=c(0,1),ylim=c(0,1),
            label="Original Risk Score",
            label2 = "Standard Recalibrated Risk Score",
            label3 = "Weighted/Constrained Recalibrated Risk Score",
            legendLab = c("Orig.", "Std.", "Wt."),
            mainTitle="Calibration of Risk Score",
            hist=TRUE,ylimHist = c(0,0.5),
            r=0.3,rl = -Inf, ru = Inf)

## End(Not run)