predict - bossanova

Prediction utilities with NA handling.

Functions:

Name	Description
`fill_valid`	Fill valid positions in result array with computed values.
`get_valid_rows`	Identify valid (non-NA) rows in a design matrix.
`init_na_array`	Create an array of NaN values.

Functions¶

fill_valid¶

fill_valid(result: NDArray, valid_mask: NDArray[np.bool_], values: NDArray) -> NDArray

Fill valid positions in result array with computed values.

Parameters:

Name	Type	Description	Default
`result`	`NDArray`	Array to fill (modified in place), shape (n,).	required
`valid_mask`	`NDArray[bool_]`	Boolean mask indicating valid positions, shape (n,).	required
`values`	`NDArray`	Values to insert at valid positions, shape (n_valid,).	required

Returns:

Type	Description
`NDArray`	The result array (same object, modified in place).

Examples:

>>> result = np.full(3, np.nan)
>>> valid_mask = np.array([True, False, True])
>>> values = np.array([10.0, 20.0])
>>> fill_valid(result, valid_mask, values)
array([10., nan, 20.])

get_valid_rows¶

get_valid_rows(X: NDArray) -> tuple[NDArray[np.bool_], NDArray, int]

Identify valid (non-NA) rows in a design matrix.

Parameters:

Name	Type	Description	Default
`X`	`NDArray`	Design matrix of shape (n, p), may contain NaN values.	required

Returns:

Type	Description
`NDArray[bool_]`	Tuple of (valid_mask, X_valid, n) where:
`NDArray`	- valid_mask: Boolean array of shape (n,), True for rows without NaN
`int`	- X_valid: Design matrix with only valid rows, shape (n_valid, p)
`tuple[NDArray[bool_], NDArray, int]`	- n: Total number of rows

Examples:

>>> X = np.array([[1, 2], [np.nan, 3], [4, 5]])
>>> valid_mask, X_valid, n = get_valid_rows(X)
>>> valid_mask
array([ True, False,  True])
>>> X_valid
array([[1., 2.], [4., 5.]])

init_na_array¶

init_na_array(n: int, dtype: type = np.float64) -> NDArray

Create an array of NaN values.

Parameters:

Name	Type	Description	Default
`n`	`int`	Length of array.	required
`dtype`	`type`	Data type (default float64).	`float64`

Returns:

Type	Description
`NDArray`	Array of NaN values with shape (n,).