BornAgain ≻ git-develop ≻ Documentation ≻ Reference ≻ Fitting ≻ Multiple datasets

Fitting several datasets at once

Several datasets can be fitted simultaneously, either using the BornAgain fit facilities, or using external minimizers.

With the BornAgain fit facilities

Add different simulation builders and data to an instance of FitObjective:

    fit_objective = ba.FitObjective()
    fit_objective.addFitPair(sim_builder1, data1[, weight1])
    fit_objective.addFitPair(sim_builder2, data2[, weight2])

Each sim_builder is a callable that takes a parameter dict and returns a Simulation object. The optional weights default to 1.

For a full example, see Examples > Fitting > Simultaneous fit.

With external minimizer

See the honeycomb fit example, where differential evolution is called with an objective function that has contributions from several reflectometry datasets and models:

    def objective_function(*args):
        ...
        return <sum of squared relative differences>

    result = scipy.optimize.differential_evolution(
        objective_function,
        ...)