Skip to main content

kernels.linearclip_1d

kernels.linearclip_1d(
    values,
    *,
    mask=None,
    low_scale=1.0,
    low=0.0,
    upp_scale=1.0,
    upp=0.0,
    maxiters=1,
    nkeep=1,
    maxrej=None,
    cenfunc='median',
    revert_on_nkeep=True,
    validate=True,
)

Center-relative linear clipping (1-D variant).

Parameters

values : (ndarray, shape(N))

One-dimensional value vector. Accepted dtypes are uint8, uint16, int16, int32, float32, and float64. Integer inputs are promoted to the package’s floating workspace when validate is True.

mask : ndarray of bool = None

Input mask; True means already masked. Must have shape (N,).

low_scale : float = 1.0

Multiplier for the per-pixel center used in the lower retained bound.

low : float = 0.0

Non-negative margin subtracted from low_scale * center.

upp_scale : float = 1.0

Multiplier for the per-pixel center used in the upper retained bound.

upp : float = 0.0

Non-negative margin added to upp_scale * center.

maxiters : int = 1

Maximum number of clipping iterations per pixel. 1 performs one clipping pass.

nkeep : int = 1

Minimum number of unmasked values to preserve at each pixel.

maxrej : int or None = None

Maximum number of values that may be rejected at each pixel. None means no limit.

cenfunc : {'median', 'lmedian', 'mean'} = 'median'

Center estimator used each iteration. lmedian accepts the alias lmed and uses the lower of the two middle samples for even valid counts.

revert_on_nkeep : bool = True

If True, an iteration that would leave fewer than nkeep usable samples at a pixel is reverted in full for that pixel.

validate : bool = True

If True, check dimensionality and normalize dtype/contiguity before entering the Rust kernel. If False, callers must provide inputs that satisfy the compiled kernel assumptions.

Returns

mask_rej : ndarray of bool, shape (N,)

True for excluded samples, including input masks and non-finite values. LinearClip excludes input-mask entries from this returned mask.

std : scalar or None

Clipping spread for SigClip; reference noise for CCDClip (see Notes). None for rejection algorithms without a spread diagnostic.

low : scalar

Minimum retained value. LinearClip instead reports its last accepted lower interval bound. Use mask_rej to identify survivors.

upp : scalar

Maximum retained value. LinearClip instead reports its last accepted upper interval bound. Use mask_rej to identify survivors.

nit : scalar

Iteration count.

output_flags : scalar

Bit-coded status using the same bit meanings as the stack kernel.

Notes

Compute c with cenfunc from the current finite, unmasked values. Retain values within the inclusive interval::

low_scale * c - low <= x <= upp_scale * c + upp

For example, low_scale=upp_scale=1, low=2, upp=3 retains values from c - 2 through c + 3. Recompute c after each accepted pass until no new samples are rejected, a safeguard reverts the pass, or maxiters is reached. maxiters=1 makes a single pass. No spread is estimated.

Default behavior differs by entry point: LinearClip() and stack linearclip() explicitly disable clipping for (low_scale, low, upp_scale, upp) = (1, 0, 1, 0); they return an all-false rejection mask, NaN bounds, and zero iterations. linearclip_1d() applies the equation even for these defaults, retaining only values equal to c unless a safeguard reverts the pass. low/upp outputs are the last accepted interval bounds, or NaN if no pass was accepted. For derivations and worked examples, see the Rejection Methods tutorial.