nlregress: Flexible nonlinear-shape regression

In R4VN: Health Data Analysis and Publication-Ready Reporting

Flexible nonlinear-shape regression

Description

Fits flexible nonlinear predictor shapes without requiring users to construct spline bases manually.

Usage

nlregress(y,
  x,
  covariates = NULL,
  data = NULL,
  spline = c("natural",
    "bspline",
    "polynomial",
    "linear"),
  df = 4,
  degree = 3,
  knots = NULL,
  boundary_knots = NULL,
  family = c("gaussian",
    "binomial",
    "poisson"),
  event = NULL,
  robust = FALSE,
  diagnosis = FALSE,
  level = 0.95,
  digits = 3,
  p_digits = 3,
  show = TRUE,
  console = FALSE)

Arguments

y	Outcome variable.
x	Numeric predictor whose functional form is modeled flexibly.
covariates	Optional additional covariates, including vars(...).
data	Data frame or active data.
spline	Natural spline, B-spline, raw polynomial, or linear form.
df	Spline degrees of freedom when knots are not supplied.
degree	B-spline/polynomial degree.
knots	Optional internal knots.
boundary_knots	Optional two boundary knots.
family	Gaussian, binomial, or Poisson model.
event	Event category for binary binomial/Poisson outcomes.
robust	Request robust covariance when available.
diagnosis	Logical; if TRUE, append diagnostics appropriate to the selected family: linear-model diagnostics for Gaussian models, calibration/influence diagnostics for binomial models, and dispersion/goodness-of-fit diagnostics for Poisson models. Default FALSE.
level, digits, p_digits, show, console	Confidence, formatting and display controls.

Details

The fitted model is stored as the active model and can be used immediately by margins(), predict() and lincom().

Value

An R4VN result object, invisibly.

See Also

qregress, margins

Examples

d <- data.frame(
  age = seq(20, 75, by = 5),
  bmi = c(20, 21, 22, 24, 23, 25, 26, 27, 29, 28, 30, 31),
  sex = factor(rep(c("Female", "Male"), 6)),
  y = c(48, 52, 55, 61, 60, 66, 69, 73, 78, 80, 85, 89),
  outcome = c(0, 0, 0, 1, 0, 1, 0, 1, 1, 0, 1, 1)
)
nlregress(y, age, data = d, spline = "natural", df = 3, show = FALSE)
nlregress(y, age, covariates = vars(sex, bmi), data = d,
          spline = "natural", knots = c(35, 50), show = FALSE)
nlregress(outcome, age, data = d, family = "binomial", event = 1,
          show = FALSE)
nlregress(y, age, data = d, spline = "bspline", df = 4, diagnosis = TRUE, show = FALSE)
nlregress(y, age, data = d, spline = "polynomial", degree = 2, show = FALSE)

R4VN documentation built on Sept. 30, 2026, 5:13 p.m.