import dascore as dc
array = dc.read_array("examples://terra15_das_1_trimmed.hdf5")
window = dc.read_array("examples://terra15_das_1_trimmed.hdf5", ((0, 10),))
window.shape == (10, array.shape[1])Trueread_array
read_array(
path: str | pathlib._local.Path | upath.core.UPath | dascore.utils.io.IOResourceManager[str, Path, UPath, IOResourceManager] ,
windows: collections.abc.Sequence[collections.abc.Sequence[Any], <class ’slice] = (),
key: str = ““,
file_format: str | None[str, None] = None,
file_version: str | None[str, None] = None,
)-> ‘np.ndarray’
Read an array, or part of one, out of a fiber file.
Parameters
| Parameter | Description |
|---|---|
| path | A path to the file to read. |
| windows |
A half-open (start, stop) sample range, a slice, or None fora whole axis, one per axis of the array in order; trailing axes left out are read whole. A bare slice is the first axis’s window. A slice may step, in which case the enclosing contiguous range is read and strided afterwards. |
| key |
Which array to read: the logical patch of a multi-patch resource, or an absolute path (“/…”) to an array stored in an HDF5 file. |
| file_format |
A string indicating the file format. If not provided dascore will try to estimate the format. |
| file_version | An optional string indicating the format version. |
Note
The array has the axis order, shape and dtype scan reports; dimension names play no part here.
An absolute HDF5 dataset path returns the stored values as they are: no reader runs, so nothing is scaled, reshaped or trimmed. OptoDAS then gives unscaled int16, and Terra15 includes its unwritten trailing rows. A single-patch format ignores a key which is not such a path.