property Data.standard_deviation#

Return data uncertainties; i.e. the estimates of the standard deviations of the noise.

If the standard_deviation property has already been set, it is returned directly. However if standard_deviations is None, then uncertainties are computed and returned according to the values assigned to the noise_floor and relative_error properties.

Where \(\varepsilon_{floor}\) is the floor uncertainty and \(C_{err}\) is the relative error for each datum \(d\), the total uncertainty \(\varepsilon\) is given by:

\[\varepsilon = \sqrt{\varepsilon_{floor}^2 + \big ( C_{err} |d| \big )^2}\]

The uncertainties applied to the data. Assuming the noise on the data are independent Gaussian with zero mean, the uncertainties represent estimates of the standard deviations of the noise on the data.