`FitMeasurement`

class nexus.FitMeasurement(*args, **kwargs)

Bases: Measurement

Abstract class for all fittable measurement class objects. Do not initialize this method, use the derived classes.

Parameters:

id (string) – String identifier.
experiment (Experiment) – Experiment.
detuning (list) – Detuning values for nuclear calculation.
time (list) – Time values for nuclear calculation which depend on time.
distribution_points (int) – Number of distribution_points for objects in forward geometry.
data_size (int) – Number of data points of experimental data.
scaling (Var) – Intensity scaling factor for fitting.
background (Var) – Background for fitting.
offset (Var) –
Offset in the x axis.

Added in version 1.1.0.
fit_weight (float) – Relative weight for the cost function in multi measurement fitting. Default is 1.
resolution (Var) – FWHM of the convolution kernel.
resolution_2 (Var) – FWHM of the convolution kernel along the second axis in 2D data sets.
kernel_type (string) – The kernel used for the convolution, either Gauss or Lorentz.

data_size

Number of data points of experimental data.

Type:: int

scaling

Intensity scaling factor for fitting.

Type:: Var

background

Background for fitting.

Type:: Var

offset

Offset in the x axis.

Added in version 1.1.0.

Type:: Var

fit_weight

Relative weight for the cost function in multi measurement fitting. Default is 1.

Type:: float

resolution

FWHM of the convolution kernel.

Type:: Var

resolution_2

FWHM of the convolution kernel along the second axis in 2D data sets.

Type:: Var

kernel_type

The kernel used for the convolution, either “Gauss” or “Lorentz”.

Type:: string

squared_residuals

Sum of the squared residuals. Calculated during fitting.

Removed in version 2.0.0.

Type:: float

sum_user_residuals

Sum of the residuals provided by the user (with exponent used). Calculated during fitting.

Added in version 2.0.0.

Type:: float

fit_residuals

The residuals from the Residual function used during fitting fit. Calculated during fitting.

Added in version 1.0.3.

Type:: array

bare_residuals

The difference residuals (:math:`x_i - y_i) from the fit. Calculated during fitting.

Added in version 1.1.0.

Type:: array

Cost()

Get the cost function of the measurement

\[cost = \frac{1}{n} \sum_i r(y_i, \hat{y}_i)\]

Added in version 2.0.0.

Returns:: cost function value.
Return type:: float

MeanSquaredError()

Get the mean squared error (MSE) of the measurement

\[cost = \frac{1}{n} \sum_i (y_i - \hat{y}_i)^2\]

Added in version 2.0.0.

Returns:: MSE value.
Return type:: float

class nexus.FitMeasurementVector(*args)

Bases: object

List of FitMeasurement objects.

FitMeasurement

`FitMeasurement`