Landmarking (version 1.0.0)

plot.landmark: Create a calibration plot

Description

Creates a calibration plot for the landmark model fitted by fit_LME_landmark_model or fit_LOCF_landmark_model. This function plots the observed frequencies of the event of interest against the predicted probabilities of the event of interest.

Usage

# S3 method for landmark
plot(x, x_L, n, x_lims, y_lims, ...)

Arguments

x

Object inheriting the class landmark, this should be the output from either fit_LME_landmark_model or fit_LOCF_landmark_model. It should contain a list of landmark models corresponding to different landmark times x_L.

x_L

Numeric specifying the landmark time. This indicates which landmark model in x to use.

n

Numeric specifying the number of bins to use.

x_lims

Vector of length 2 specifying the limits of the x axes

y_lims

Vector of length 2 specifying the limits of the y axes

…

Arguments passed to ggplot2::labs to modify axis, legend, and plot labels

Value

Calibration plot showing the value of predicted probabilities against observed frequencies, with a y=x line.

Details

This function bins the predicted probabilities of the event of interest into n bins. The event of interest is the event with event_status=1 when fitting the landmark model. For each of the n sets of individuals, the Aalen-Johansen estimator is fit to that set and used to calculate the risk of an event at the horizon time. The predictions (from the landmark model) and the observed frequencies (from the Aalen-Johansen estimator) are plotted against each other. For a perfect prediction model, the points will be plotted along the y=x line.

Examples

# NOT RUN {
library(Landmarking)
data(data_repeat_outcomes)
data_model_landmark_LOCF <-
  fit_LOCF_landmark(
    data_long = data_repeat_outcomes,
    x_L = c(60, 61),
    x_hor = c(65, 66),
    covariates =
      c("ethnicity", "smoking", "diabetes", "sbp_stnd", "tchdl_stnd"),
    covariates_time =
      c(rep("response_time_sbp_stnd", 4), "response_time_tchdl_stnd"),
    k = 10,
    individual_id = "id",
    event_time = "event_time",
    event_status = "event_status",
    survival_submodel = "cause_specific"
  )
 plot(x=data_model_landmark_LOCF,x_L=60,n=5)
 plot(x=data_model_landmark_LOCF,x_L=61,n=5)
# }