SChunk.fill_special#

SChunk.fill_special(nitems: int, special_value: SpecialValue, value: bytes | int | float | bool = None) → int#

Fill the SChunk with a special value. SChunk must be empty.

Parameters:

  • nitems (int) – The number of items to fill with the special value.

  • special_value (SpecialValue) – The special value to be used for filling the SChunk.

  • value (bytes, int, float, bool (optional)) – The value to fill the SChunk. This parameter is only supported if special_value is blosc2.SpecialValue.VALUE.

Returns:

out – The number of chunks in the SChunk.

Return type:

int

Raises:

RunTimeError – If the SChunk could not be filled with the special value.

Examples

>>> import blosc2
>>> import numpy as np
>>> import time
>>> nitems = 100_000_000
>>> dtype = np.dtype(np.float64)
>>> # Measure the time to create SChunk from a NumPy array
>>> t0 = time.time()
>>> data = np.full(nitems, np.pi, dtype)
>>> schunk = blosc2.SChunk(data=data, cparams={"typesize": dtype.itemsize})
>>> t = (time.time() - t0) * 1000.
>>> f"Time creating a schunk with a numpy array: {t:10.3f} ms"
Time creating a schunk with a numpy array:    710.273 ms
>>> # Measure the time to create SChunk using fill_special
>>> t0 = time.time()
>>> schunk = blosc2.SChunk(cparams={"typesize": dtype.itemsize})
>>> schunk.fill_special(nitems, blosc2.SpecialValue.VALUE, np.pi)
>>> t = (time.time() - t0) * 1000.
>>> f"Time passing directly the value to `fill_special`: {t:10.3f} ms"
Time passing directly the value to `fill_special`:      2.109 ms
On this page