Function Backgrounds

RAT supports function backgrounds, which use a custom function to describe the background for the experiment.

The signature for background functions is shown in the code snippet below. The first argument xdata is a vector containing the q points over the simulation range, which should incorporate the q values of the supplied data and the second argument params is a vector containing the background parameters associated with the background, up to 5 can be related to a specific function background. The function should return an array with the background value for each of the input simulation points.

MATLAB

function background = backgroundFunction(xdata, params)

    background = zeros(numel(xdata), 1);

end

PYTHON

import numpy as np


def background_function(xdata, params):

    background = np.zeros(len(xdata))
    return background

To define a function background, add the custom function, associated background parameters, then create the background as shown below

MATLAB

% Add the function background
% Add the function to custom files, and its parameters to background
% parameters
problem.addCustomFile('Back Fun','backgroundFunction.m','matlab',pwd);

problem.addBackgroundParam('Fn Ao',5e-7, 8e-6, 5e-5);
problem.addBackgroundParam('Fn k',40, 70, 90);
problem.addBackgroundParam('Fn Const',1e-7, 8e-6, 1e-5);

% Add the Background
problem.addBackground('Func Background','function','Back Fun','Fn Ao','Fn k','Fn Const');

PYTHON

problem.custom_files.append(name="Back Fun", filename="background_function.py", language="python")

problem.background_parameters.append(name="Fn Ao", min=5e-7, value=8e-6, max=5e-5)
problem.background_parameters.append(name="Fn k", min=40, value=70, max=90)
problem.background_parameters.append(name="Fn Const", min=1e-7, value=8e-6, max=1e-5)

problem.backgrounds.append(name="Func Background", type="function", source="Back Func", value_1="Fn Ao", value_2="Fn k", value_3="Fn Const")

An example of using Python and Matlab custom models can be found in examples folder. Look at DSPCScriptWithFunctionBackground.m and DSPC_function_background.py in Matlab and Python respectively.