keras-team/keras · error · ValueError

Invalid value for argument `method`. Expected of one

Error message

Invalid value for argument `method`. Expected of one {SCALE_AND_TRANSLATE_METHODS}. Received: method={method}

What it means

Argument validation at the top of `scale_and_translate` (NumPy backend): the resampling `method` is checked against the SCALE_AND_TRANSLATE_METHODS set ('linear', 'bilinear', 'trilinear', 'cubic', 'bicubic', 'tricubic', 'lanczos3', 'lanczos5'). Any other string — wrong casing, a typo, or a method supported only by `resize` — fails the membership test and is rejected with the allowed set listed.

Solutions

1. Pass one of the supported method strings exactly as printed (e.g. 'bilinear', 'lanczos3')
2. Map resize-style names to their equivalents here ('bicubic' not 'cubic' if 'cubic' is rejected by your version's set)
3. Validate against the printed set and fall back to 'bilinear' for unknown user-supplied methods

When it happens

Trigger: Thrown at keras/src/backend/numpy/ops/image.py:1189 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of keras-team/keras@7a34a03db6 (2026-08-25). Data as JSON: /api/errors/42b9343405592f1c. Report an issue: GitHub.

Appendix: source

Thrown at keras/src/backend/numpy/ops/image.py:1189

    if need_squeeze:
        transformed_images = np.squeeze(transformed_images, axis=0)
    transformed_images = transformed_images.astype(input_dtype)

    return transformed_images


def scale_and_translate(
    images,
    output_shape,
    scale,
    translation,
    spatial_dims,
    method,
    antialias=True,
):
    if method not in SCALE_AND_TRANSLATE_METHODS:
        raise ValueError(
            "Invalid value for argument `method`. Expected of one "
            f"{SCALE_AND_TRANSLATE_METHODS}. Received: method={method}"
        )
    if method in ("linear", "bilinear", "trilinear", "triangle"):
        method = "linear"
    elif method in ("cubic", "bicubic", "tricubic"):
        method = "cubic"

    images = convert_to_tensor(images)
    scale = convert_to_tensor(scale)
    translation = convert_to_tensor(translation)
    kernel = _kernels[method]
    dtype = backend.result_type(scale.dtype, translation.dtype)
    scale = scale.astype(dtype)
    translation = translation.astype(dtype)
    return _scale_and_translate(
        images,
        output_shape,

View on GitHub (pinned to 7a34a03db6)