map

map(
    self ,
    func: collections.abc.Callable[…, None] ,
    client: dascore.constants.ExecutorType | None[ExecutorType, None] = None,
    size: int | None[int, None] = None,
    progress: Literal[‘standard’, ‘basic’, None] = standard,
    **kwargs ,
)-> ‘list[T]’

Map a function of all the contents of the spool.

Parameters

Parameter	Description
func	A callable which takes a patch as its first argument.
client	A client, or executor, which has a `map` method.
size	The number of patches in each spool mapped to a client. If not set, defaults to the number of processors on the host. Does nothing unless client is defined.
progress	Controls the progress bar. “standard” produces the standard progress bar. “basic” is a simplified version with lower refresh rates, best for high-latency environments, and None disables the progress bar.
**kwargs	kwargs passed to func.

Note

When a client is specified, the spool is split then passed to the client’s map method. This is to avoid serializing loaded patches. See Spool.split for more details about the size and count parameters.

Examples

import numpy as np
import dascore as dc

spool = dc.get_example_spool("random_das")

# Calculate the std for each channel in 5 second chunks
results = (
     spool.chunk(time=5)
     .map(lambda x: np.std(x.data, axis=0))
)
# stack back into array. dims are (distance, time chunk)
out = np.stack(results, axis=-1)