# Coefficients for SVEM Models Extracts averaged coefficients from an `svem_model` fitted by `SVEMnet`. ```r ## S3 method for class 'svem_model' coef(object, debiased = FALSE, ...) ``` ## Arguments - `object`: An object of class `svem_model`, typically returned by `SVEMnet`. - `debiased`: Logical; if `TRUE` and debiased coefficients are available (Gaussian fits with `parms_debiased`), return those instead of `parms`. Default is `FALSE`. - `...`: Unused; present for S3 method compatibility. ## Returns A named numeric vector of coefficients (including the intercept). ## Details For Gaussian fits, you can optionally request debiased coefficients (if they were computed and stored) via `debiased = TRUE`. In that case, the function returns `object$parms_debiased`. If debiased coefficients are not available, or if `debiased = FALSE`, the function returns `object$parms`, which are the ensemble-averaged coefficients across bootstrap members. For Binomial models, `debiased` is ignored and the averaged coefficients in `object$parms` are returned. This is a lightweight accessor around the stored components of an `svem_model`: * `parms`: ensemble-averaged coefficients over bootstrap members, on the model's link scale; * `parms_debiased`: optional debiased coefficients (Gaussian only), if requested at fit time. Passing `debiased = TRUE` has no effect if `parms_debiased` is `NULL`. ## Examples ```r set.seed(1) n <- 200 x1 <- rnorm(n) x2 <- rnorm(n) eps <- rnorm(n, sd = 0.3) y_g <- 1 + 2*x1 - 0.5*x2 + eps dat_g <- data.frame(y_g, x1, x2) # Small nBoot to keep runtime light in examples fit_g <- SVEMnet(y_g ~ x1 + x2, data = dat_g, nBoot = 30, relaxed = TRUE) # Ensemble-averaged coefficients cc <- coef(fit_g) head(cc) # Debiased (only if available for Gaussian fits) ccd <- coef(fit_g, debiased = TRUE) head(ccd) # Binomial example (0/1 outcome) set.seed(2) n <- 250 x1 <- rnorm(n) x2 <- rnorm(n) eta <- -0.4 + 1.1*x1 - 0.7*x2 p <- 1/(1+exp(-eta)) y_b <- rbinom(n, 1, p) dat_b <- data.frame(y_b, x1, x2) fit_b <- SVEMnet(y_b ~ x1 + x2, data = dat_b, family = "binomial", nBoot = 30, relaxed = TRUE) # Averaged coefficients (binomial; debiased is ignored) coef(fit_b) ``` ## See Also `svem_nonzero` for bootstrap nonzero percentages and a quick stability plot.

coef.svem_model function