smash.Model.bayes_estimate

Model.bayes_estimate(sample=None, alpha=4, n=1000, random_state=None, jobs_fun='nse', wjobs_fun=None, event_seg=None, gauge='downstream', wgauge='mean', ost=None, ncpu=1, verbose=True, inplace=False, return_br=False)

Estimate prior Model parameters/states using Bayesian approach.

Hint

See the User Guide and Math / Num Documentation for more.

Parameters:

sampleSampleResult or None, default None

An instance of the SampleResult object, which should be created using the smash.generate_samples method.

Note

If not given, the Model parameters samples will be generated automatically using the uniform generator based on the Model structure considered.

alphaint, float or sequence, default 4

A regularization parameter that controls the decay rate of the likelihood function.

Note

If alpha is a sequence, then the L-curve approach will be used to find an optimal value for the regularization parameter.

nint, default 1000

Number of generated samples. Only used if sample is not set.

random_stateint or None, default None

Random seed used to generate samples. Only used if sample is not set.

Note

If not given and sample is not set, generates the parameters set with a random seed.

jobs_fun, wjobs_fun, event_seg, gauge, wgauge, ostmultiple types

Optimization setting to run the forward hydrological model and compute the cost values. See smash.Model.optimize for more.

ncpuinteger, default 1

If ncpu > 1, perform a parallel computation through the parameters set.

verbosebool, default True

Display information while estimating.

inplacebool, default False

If True, perform operation in-place.

return_brbool, default False

If True, also return the Bayesian estimation result BayesResult.

Returns:

ModelModel or None

Model with optimize outputs if not inplace.

resBayesResult

The Bayesian estimation results represented as a BayesResult object if return_br.

See also

BayesResult

Represents the Bayesian estimation or optimization result.

SampleResult

Represents the generated sample result.

Examples

>>> setup, mesh = smash.load_dataset("cance")
>>> model = smash.Model(setup, mesh)
>>> br = model.bayes_estimate(n=200, inplace=True, return_br=True, random_state=99)

Access to cost values of the direct simulations

>>> cost = br.data["cost"]
>>> cost.sort()  # sort the values by ascending order
>>> cost
array([4.88620669e-01, 5.70850313e-01, 7.37333179e-01, 7.59980202e-01,
    ...
    9.26341797e+03, 1.12409111e+04, 1.13607480e+04, 1.18410371e+04])

Compare to the cost value of the Model with the estimated parameters using Bayesian apporach

>>> model.output.cost
0.41908782720565796