Utility Functions

Contains utility functions and classes used in library.

spiegelib.core.utils.spectrum_types = ['complex', 'magnitude', 'power', 'magnitude_phase', 'power_phase']

Selection of spectrum outputs

spiegelib.core.utils.convert_spectrum(spectrum, type, dtype=<class 'numpy.float32'>, complex_dtype=<class 'numpy.complex64'>)

Convert a raw spectrum to magnitue, power, magnitude with phase, or power with phase. If output is magnitude or power, the output array shape is the same as the input. If either of the options with phase are selected, the output is the same shape but each element is replaced with a 2-element vector containing the magnitude or power in the first element and the phase in the second.

Parameters

  • spectrum (np.ndarray) – array of complex values

  • type (str) – type of conversion to apply (‘magnitude’, ‘power’, ‘magnitude_phase’, ‘power_phase’)

  • dtype (np number type, optional) – number type of regular (non-complex) numbers to return, defaults to np.float32

  • complex_dtype (np complex type, optional) – complex number type, defaults to np.complex64

Returns

converted spectrum

Return type

np.ndarray

spiegelib.core.utils.atoi(text)

Convert string to integer if it is an integer, otherwise keep it a str

Parameters

text (str) –

Returns

int or str

spiegelib.core.utils.natural_keys(text)

Keys for sorting text in natural ordering. For text containing numbers, makes sure numbers are sorted in a natural way.

Usage:

alist.sort(key=natural_keys) sorts in human order

From stackoverflow

Parameters

text (str) – text to extract natural keys from

Returns

List of text and integers separated from text to use as the key argument in a sort algorithm.

Return type

list

class spiegelib.core.utils.NumpyNumberEncoder(*, skipkeys=False, ensure_ascii=True, check_circular=True, allow_nan=True, sort_keys=False, indent=None, separators=None, default=None)

Encodes Numpy number values into values that can be saved as JSON files.

Usage:

with open(path, 'w') as fp:
    json.dump(data, fp, indent=True, cls=NumpyNumberEncoder)