read_array

read_array(
    self ,
    resource ,
    windows: dict[str, tuple[int, int]] ,
    **kwargs ,
)-> ‘np.ndarray’

Return the raw data array for absolute sample windows.

A data-only fast path for callers which already know a resource’s structure (from an index) and need no Patch, attrs, or coordinates back. This default reads the whole resource through read and trims — a Patch is still built internally, just not returned — so every format is correct without overriding; formats override it to slice storage directly and skip the Patch work.

Parameters

Parameter	Description
resource	The resource to read, as `read` takes it.
windows	Maps dimension name to `(start, stop)` half-open python indices on the resource’s own sample grid. Dimensions absent from the mapping are returned whole.
**kwargs	Reader-specific options. Multi-patch resources take `source_patch_key` (as `read` and `scan` spell it) naming the one patch the windows index; without it an ambiguous resource raises rather than guesses.

Returns

The array in the resource’s stated dimension order (the order scan reports), untransposed and uncast.

Examples

from dascore.io.dasdae.core import DASDAEV1
from dascore.utils.downloader import fetch

path = fetch("example_dasdae_event_1.h5")
array = DASDAEV1().read_array(path, {"time": (0, 50)})
array.shape

(601, 50)