Predict treatment contrasts from a fitted network_meta model

Usage

# S3 method for class 'nma_mars'
predict(
  object,
  newdata,
  se.fit = FALSE,
  interval = c("none", "confidence", "prediction"),
  level = 0.95,
  treatment_1 = "treatment_1",
  treatment_2 = "treatment_2",
  ...
)

Arguments

object: A fitted nma_mars object.
newdata: Data frame containing treatment_1 and treatment_2 columns for contrasts to predict.
se.fit: Logical; if TRUE, returns an approximate prediction standard error.
interval: Type of interval; one of "none", "confidence", or "prediction". Prediction intervals add the fitted model's between-study variance estimate to the contrast variance.
level: Confidence level for intervals.
treatment_1, treatment_2: Column names used in newdata.
...: Unused.

Value

Data frame with predicted contrast estimates and optional uncertainty.