Grid

class of dascore.core.coords

source

Grid(
    origin: Any ,
    step_num: Any ,
    step_den: int ,
    count: int ,
    k0: int = 0,
    phase: int = 0,
    parent_count: int | None[int, None] = None,
    stride: int = 1,
)-> None

One evenly sampled run of labels.

An exact run (step_den above zero) works in ticks – nanoseconds for time, the integers themselves otherwise – and labels sample k as origin + (phase + (k0 + k) * step_num) // step_den. The ideal grid therefore has a spacing of step_num / step_den ticks and an origin phase / step_den ticks after origin, and each label is the floor of its ideal position, so a 1024 Hz grid never drifts.

A float run (step_den of zero) keeps step_num as a scalar step and spaces its labels linearly, as numpy does. A slice keeps the original parent_count, origin and step, and maps samples through the integer offset k0 and stride. Evaluating the original expression keeps every label bit-for-bit, including numpy’s endpoint rounding.

An exact run folds k0 into its origin and phase. A newly constructed float run also folds an explicit offset into its origin; only slices with parent_count retain their original grid.

Methods

Name	Description
canonical	The terms which name the run’s labels: origin, step, and phase.
index_of	The sample index each label tick maps to.
labels	The labels at these indices, which may lie outside the run.
resized	Restate a grid’s extent, as an explicit resize or gap fill requests.
sliced	The run of the samples first, first + stride, … (count of them).
step	The whole-tick (or float) spacing between neighbouring labels.
step_exact	The exact spacing in coordinate units, seconds for time.