Coxmos (version 1.1.2)

plot_observation.eventHistogram: plot_observation.eventHistogram

Description

Generates a histogram plot for observation event data based on a given Coxmos model. The function visualizes the distribution of predicted values and highlights the prediction for a specific observation.

Usage

plot_observation.eventHistogram(
  observation,
  model,
  time = NULL,
  type = "lp",
  size = 3,
  color = "red"
)

Value

A ggplot object representing a histogram of the predicted event values based on the provided Coxmos model.

Arguments

observation

Numeric matrix or data.frame. New explanatory variables (raw data) for one observation. Qualitative variables must be transform into binary variables.

model

Coxmos model.

time

Numeric. Time point where the AUC will be evaluated (default: NULL).

type

Character. Prediction type: "lp", "risk", "expected" or "survival" (default: "lp").

size

Numeric. Point size (default: 3).

color

String. R Color.

Author

Pedro Salguero Garcia. Maintainer: [e-mail]

Details

The plot_observation.eventHistogram function is designed to provide a visual representation of the distribution of predicted event values based on a Coxmos model. The function takes in observation data, a specified time point, and a Coxmos model to compute the prediction. The resulting histogram plot displays the distribution of these predictions, with a specific emphasis on the prediction for the provided observation data. The prediction is represented as a point on the histogram, allowing for easy comparison between the specific observation's prediction and the overall distribution of predictions. The type of prediction ("lp", "risk", "expected", or "survival") can be specified, offering flexibility in the kind of insights one wishes to derive from the visualization. The appearance of the point representing the observation's prediction can be customized using the size and color parameters.

Examples

data("X_proteomic")
data("Y_proteomic")
set.seed(123)
index_train <- caret::createDataPartition(Y_proteomic$event, p = .5, list = FALSE, times = 1)
X_train <- X_proteomic[index_train,1:50]
Y_train <- Y_proteomic[index_train,]
X_test <- X_proteomic[-index_train,1:50]
Y_test <- Y_proteomic[-index_train,]
coxEN.model <- coxEN(X_train, Y_train, x.center = TRUE, x.scale = TRUE)
observation = X_test[1,,drop=FALSE]
plot_observation.eventHistogram(observation = observation, model = coxEN.model, time = NULL)