S3 plot method for objects of class "glow_B_model" returned by
train_B_model. Provides model-level diagnostic plots
including model fit, residuals, Q-Q plots, and model comparison.
Unlike plot.glow_B_estimate, this method works without
target/prediction data. It focuses on training-time model diagnostics.
# S3 method for class 'glow_B_model'
plot(x, type = "all", which_method = "primary", ...)Object of class "glow_B_model" from train_B_model
Character string specifying plot type:
"all" (default): 2x2 dashboard with model diagnostics
"fit": Model fit plot showing training data vs fitted curve
"residuals": Standard residual diagnostic plots (Q-Q, residuals vs fitted)
"models": Compare all 8 candidate model fits (bar chart)
Character string, which method to plot: "beta", "pvalue", or "primary" (default, uses the primary/selected method)
Additional arguments passed to plotting functions
NULL (invisibly). Plots are displayed as a side effect.
This method produces model-level diagnostics only (no prediction-based
plots, since a glow_B_model does not contain B predictions).
Dashboard View (type = "all"):
Top-left: Fitted model vs training data
Top-right: Normal Q-Q plot of residuals
Bottom-left: Residuals vs fitted values
Bottom-right: Cook's distance
Comparison Note:
The "comparison" type from plot.glow_B_estimate is not
available here because it requires predicted B vectors. Use
type = "models" to compare all candidate model fits instead.
train_B_model for training B estimation models
plot.glow_B_estimate for prediction-level diagnostics
if (FALSE) { # \dontrun{
b_model <- train_B_model(
training_trait = "binary",
training_MAF = training_MAF,
training_BETA = training_BETA,
method = "beta"
)
# Dashboard view
plot(b_model)
# Individual plots
plot(b_model, type = "fit")
plot(b_model, type = "residuals")
plot(b_model, type = "models")
} # }