hat_values-fit.md
In applicable: A Compilation of Applicability Domain Methods

XtX_inv is provided

Code
  new_apd_hat_values(blueprint = hardhat::default_xy_blueprint())
Condition
  Error in `new_apd_hat_values()`:
  ! argument "XtX_inv" is missing, with no default

`new_apd_hat_values` fails when blueprint is numeric

Code
  new_apd_hat_values(XtX_inv = 1, blueprint = 1)
Condition
  Error in `hardhat::new_model()`:
  ! `blueprint` must be a <hardhat_blueprint>, not the number 1.

`apd_hat_values` fails when matrix has more predictors than samples

Code
  apd_hat_values(bad_data)
Condition
  Error in `apd_hat_values_bridge()`:
  ! The number of columns must be less than the number of rows.

`apd_hat_values` fails when the matrix X^tX is singular

Code
  apd_hat_values(bad_data)
Condition
  Error in `get_inv()`:
  ! Unable to compute the hat values of the matrix X of
  predictors because the matrix resulting from multiplying
  the transpose of X by X is singular.