Validate regression model inputs¶
Source: R/validation.R ยท Help: ?rvalidateModelInputs
Ensures that a supplied regression model satisfies the basic assumptions
required by the heteroscedasticity testing infrastructure. The checks
include class validation, successful fitting, availability of finite
residuals, and minimum sample size requirements. The residuals are those of
the observations of the fit: on a fit with na.action = na.exclude, without
the NA that residuals() pads them with at the rows the fit dropped, as
for the same fit with na.omit (since 0.17.0; before, such a fit was
refused as one whose residuals are not finite).
Usage¶
Arguments¶
model- A fitted model object produced by
stats::lm()orstats::glm(). test_name- A scalar character identifier used in error messages to reference the calling test.
min_obs- Minimum number of observations required for the calling
procedure. Defaults to
10. allow_aliased- Logical.
FALSE, the default, refuses a rank-deficient fit, whose aliased coefficients areNA, with a message that names them;TRUEaccepts it, for a procedure computed on the estimable part of the design. A coefficient that is not finite for another reason is refused either way.
Value¶
Invisibly returns model when validation passes.