pymrm.ibm.apply_ibm_vector

def apply_ibm_vector(vec, ibm):
    """Apply the IBM per-row conditioning scale to a flat vector.

    When the operator matrix is constant it is modified once via
    :func:`apply_ibm`; any independent right-hand-side term must be scaled
    by the same per-row factor to keep the system consistent.  This helper
    applies the combined scale for both sides of the interface and expands
    from the spatial grid to the full field.

    Parameters
    ----------
    vec : array_like
        Vector to scale, must have ``size == ibm.n_cells``.  May be shaped
        as the full field or passed as a flat array.
    ibm : IBM
        Immersed-boundary data from :func:`construct_ibm`.

    Returns
    -------
    numpy.ndarray, shape (n_cells,)
        Row-scaled vector, always returned as a 1-D flat array.
    """
    vec = np.asarray(vec, dtype=float).ravel()
    if vec.size != ibm.n_cells:
        raise ValueError(
            f"vec size {vec.size} != n_cells {ibm.n_cells}"
        )
    out = vec.copy()
    scale_rows, scale_vals = _combined_cut_scale(ibm)
    out[scale_rows] *= scale_vals
    return out