elephant.utils.round_binning_errors

elephant.utils.round_binning_errors(values, tolerance=1e-08)

Round the input values in-place due to the machine floating point precision errors.

Parameters:

valuesnp.ndarray or float

An input array or a scalar.

tolerancefloat or None, optional

The precision error absolute tolerance; acts as atol in numpy.isclose() function. If None, no rounding is performed. Default: 1e-8

Returns:

valuesnp.ndarray or int

Corrected integer values.

Examples

>>> from elephant.utils import round_binning_errors
>>> round_binning_errors(0.999999, tolerance=None)
0
>>> round_binning_errors(0.999999, tolerance=1e-6)
1