jax-ml/jax · error · ValueError

Unrecognized {mode=}

Error message

Unrecognized {mode=}

What it means

Defensive fallthrough in _fftconvolve_unbatched when mode is not 'full'/'same'/'valid' after slicing logic. Normally unreachable because fftconvolve validates mode earlier; it can fire via direct internal calls or monkeypatching.

Source

Thrown at jax/_src/scipy/signal.py:152

  if (all(s1 == 1 or s2 == 1 for s1, s2 in zip(in1.shape, in2.shape))):
    conv = in1 * in2
  else:
    if jnp.iscomplexobj(in1):
      fft, ifft = jnp.fft.fftn, jnp.fft.ifftn
    else:
      fft, ifft = jnp.fft.rfftn, jnp.fft.irfftn
    sp1 = fft(in1, fft_shape)
    sp2 = fft(in2, fft_shape)
    conv = ifft(sp1 * sp2, fft_shape)

  if mode == "full":
    out_shape = full_shape
  elif mode == "same":
    out_shape = in1.shape
  elif mode == "valid":
    out_shape = tuple(s1 - s2 + 1 for s1, s2 in zip(in1.shape, in2.shape))
  else:
    raise ValueError(f"Unrecognized {mode=}")

  start_indices = tuple((full_size - out_size) // 2
                        for full_size, out_size in zip(full_shape, out_shape))
  return lax.dynamic_slice(conv, start_indices, out_shape)


# Note: we do not reuse the code from jax.numpy.convolve here, because the handling
# of padding differs slightly between the two implementations (particularly for
# mode='same').
def _convolve_nd(in1: Array, in2: Array, mode: ModeString, *, precision: PrecisionLike) -> Array:
  if mode not in ["full", "same", "valid"]:
    raise ValueError("mode must be one of ['full', 'same', 'valid']")
  if in1.ndim != in2.ndim:
    raise ValueError("in1 and in2 must have the same number of dimensions")
  if in1.size == 0 or in2.size == 0:
    raise ValueError(f"zero-size arrays not supported in convolutions, got shapes {in1.shape} and {in2.shape}.")
  in1, in2 = promote_dtypes_inexact(in1, in2)

View on GitHub (pinned to 1e1c6a8fc0)

Solutions

1. Call the public fftconvolve/convolve instead of private helpers
2. Validate mode against ['full','same','valid'] before any internal pass-through

Example fix

# before
out = _fftconvolve_unbatched(in1, in2, mode='circul')
# after
out = jax.scipy.signal.fftconvolve(in1, in2, mode='same')

assert mode in ('full','same','valid'), mode

def is_mode(m: str) -> bool: return m in ('full', 'same', 'valid')

When it happens

Trigger: Calling the private _fftconvolve_unbatched with an unchecked mode; bypassing the public API validation.

Common situations: Internal code reuse where mode is passed through from another function without revalidation.

Related errors

• mode must be one of ['same', 'full', 'valid']
• mode must be one of ['full', 'same', 'valid']
• Unsupported method: {method}
• ind must be a positive integer; got {ind=}
• Expected kind to be on of: {valid_kind}; got {kind}

AI-assisted analysis of jax-ml/jax@1e1c6a8fc0 (2026-08-27). Data as JSON: /api/errors/58f751992d4e176f. Report an issue: GitHub.