lehmann_roc: Lehmann model for ROC curves.

Description Usage Arguments Value Examples

View source: R/roc.R

Description

Relevant statistics for Lehmann models of ROC curves.

Usage

1

Arguments

formula

A formula containing the covariates to be analyzed.

data

A data.frame containing the variables named in the formula.

Value

The output is a list of relavant statistics. theta, var_theta, auc, and var_auc are all numerics. partial_auc and var_partial_auc are both functions of the false positive rate. roc is a list containing numeric vectors TPR and FPR. var_roc is a function of the false positive rate. If the input formula contains concomitant covariates, the return list will contain functions. For example, theta would now be a function of the covariate values. The apply_covariates function should be used to input values of covariates into appropriate lehmann_roc objects.

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
# Import a dataset
ovarian <- survival::ovarian
head(ovarian)

# Create a Lehmann ROC model using only disease status.
l <- lehmann_roc(futime~resid.ds, ovarian)
summary(l)
plot(l)
l$theta
l$auc
l$partial_auc(0.5)

# Create a Lehmann ROC model with age as a concomitant covariate.
l <- lehmann_roc(futime~resid.ds*age, ovarian)

# The output is now a list of functions
summary(l)

## Not run: 
l # errors
plot(l) # errors

## End(Not run)

# Use the apply_covariates function
l_objects <- apply_covariates(l, list(age=21))
l2 <- l_objects[["21"]]
l2
l2$partial_auc(0.5)
plot(l2)

Nejiv/Lehmann-ROC documentation built on Oct. 13, 2020, 10:52 a.m.