model_gen function

Model Generations

Generate appropriate regression models based on the model type and formula

model_gen(
  y,
  type = "lm",
  batch = NULL,
  covariates = NULL,
  interaction = NULL,
  random = NULL,
  smooth = NULL,
  df
)

Arguments

• y: Dependent variable in the model.
• type: A model function name that is used or to be used in the ComBatFamily Package (eg: "lmer", "lm", "gam").
• batch: Name of batch variable (often equivalent to site or scanner).
• covariates: Name of covariates supplied to model.
• interaction: Expression of interaction terms supplied to model (eg: "age:diagnosis").
• random: Variable name of a random effect in linear mixed effect model.
• smooth: Variable name that requires a smooth function.
• df: Dataset to be harmonized.

Returns

A regression model object to be used for batch effect diagnostics and the post-harmonization stage.

Examples

model_gen(y = "thickness.left.caudal.anterior.cingulate", type = "lm",
batch = "manufac", covariates = c("AGE", "SEX"), df = adni)