MIDASVAR._blocks

MIDASVAR._blocks()

Assemble the fixed pieces the profile search reuses at every draw.

Return type:

tuple[NDArray[double], NDArray[double], NDArray[double]]

Returns:

The target block, the deterministic-and-lag design, and the (effective, midas_lags, m) window stack.

Example

>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> x = rng.standard_normal((120, 1))
>>> model = MIDASVAR(rng.standard_normal((40, 2)), x, order=1, period=3)
>>> target, base, windows = model._blocks()
>>> target.shape, base.shape, windows.shape
((39, 2), (39, 3), (39, 3, 1))
>>> bool(np.allclose(windows[0, :, 0], x[5:2:-1, 0]))
True