Skip to content

FitUtil

Moose.FitUtil

Contains convenience functions, to streamline the workflow a bit when using lmfit as a fitting library.

Whenever lmfit is installed in the currently active environment, these functions are imported into the Moose namespace as well.

This behaviour occurs regardless of if Moose has been installed with the [fit] optional dependencies or not.

Functions:

Name Description
make_model

Create a lmfit.Model instance already prepared with the line-by-line database for fitting.
set_param

Set or modify a single parameter in lmfit.Parameters object.
set_params

Set or modify a bunch of parameters using a dict in a lmfit.Parameters object.

make_model 

make_model(species: str, range: tuple[float, float] = (0, np.inf), resolution: int = 100, wl_pad: float = 10, **kwargs)

Create a lmfit.Model instance already prepared with the line-by-line database for fitting.

Parameters:

Name Type Description Default
species str

The name of the species database to query.

 required
range tuple[float, float]

The wavelength range in nanometer to restric the database query to. Defaults to (0,np.inf)

 (0, inf)
resolution int

The resolution per nanometer of the equidistant mesh compared to bin/sample simulation by (default: 100)

 100
wl_pad float

The amount of nanometer to pad the x-axis of the simulation with to avoid edge effects. Default: 10

 10
Kwargs

params (lmfit.Parameters, Optional): A Parameters object with appropriate fit parameters for the model. Defaults to Moose.default_params

Returns:

Name Type Description
model Model

A Model object that can be used for fitting
params Parameters

A Parameters object that can be used for fitting, containing default initial values and ranges.
Source code in src/Moose/FitUtil.py 
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
def make_model(
    species: str,
    range: tuple[float, float] = (0, np.inf),
    resolution: int = 100,
    wl_pad: float = 10,
    **kwargs,
):
    """Create a `lmfit.Model` instance already prepared with the line-by-line database for fitting.

    Args:
        species (str): The name of the species database to query.
        range (tuple[float,float]): The wavelength range in nanometer to restric the database query to. Defaults to (0,np.inf)
        resolution (int):   The resolution per nanometer of  the equidistant mesh compared to bin/sample simulation by (default: 100)
        wl_pad (float): The amount of nanometer to pad the x-axis of the simulation with to avoid edge effects. Default: 10

    Kwargs:
        params (lmfit.Parameters, Optional): A `Parameters` object with appropriate fit parameters for the model. Defaults to `Moose.default_params`

    Returns:
        model (lmfit.Model): A `Model` object that can be used for fitting
        params (lmfit.Parameters): A `Parameters` object that can be used for fitting, containing default initial values and ranges.
    """
    pars = kwargs.pop("params", default_params)
    db = query_DB(species, range, **kwargs)
    # param_names = kwargs.pop('param_names', [p for p in _default_params.keys()]) # Makes it possible to change what are considered fit parameters. Be default only `wl_pad` and `resolution` are excluded
    model = lmfit.Model(model_for_fit, sim_db=db)
    params = lmfit.create_params(**pars)
    params["wl_pad"].value = wl_pad
    params["resolution"].value = resolution
    return model, params

set_param 

set_param(params: Parameters, param_name: str, value: float = 0, min: float = -np.inf, max: float = np.inf, vary: bool = True)

Set or modify a single parameter in lmfit.Parameters object.

Source code in src/Moose/FitUtil.py 
15
16
17
18
19
20
21
22
23
24
def set_param(
    params: lmfit.Parameters,
    param_name: str,
    value: float = 0,
    min: float = -np.inf,
    max: float = np.inf,
    vary: bool = True,
):
    """Set or modify a single parameter in `lmfit.Parameters` object."""
    params[param_name].set(value=value, min=min, max=max, vary=vary)

set_params 

set_params(params: Parameters, param_dict: dict = default_params, print: bool = False)

Set or modify a bunch of parameters using a dict in a lmfit.Parameters object.

Source code in src/Moose/FitUtil.py 
27
28
29
30
31
32
def set_params(params: lmfit.Parameters, param_dict: dict = default_params, print: bool = False):
    """Set or modify a bunch of parameters using a dict in a `lmfit.Parameters` object."""
    for param in param_dict:
        set_param(params, param, **param_dict[param])
    if print:
        params.pretty_print()