WorkTrackingModel

class pyapprox.interface.WorkTrackingModel(function, base_model=None, num_config_vars=0, enforce_timer_model=True)[source]

Bases: object

Keep track of the wall time needed to evaluate a function.

Methods Summary

__call__(samples[, return_grad])

Evaluate self.function

cost_function([config_samples])

Retrun the cost of evaluating the functions with the ids given in a set of config_samples.

Methods Documentation

__call__(samples, return_grad=False)[source]

Evaluate self.function

Parameters:
samplesnp.ndarray (nvars,nsamples)

Samples used to evaluate self.function

Returns:
valuesnp.ndarray (nsamples,nqoi)

The values of self.function. The last qoi returned by self.function (i.e. the last column of the output array of size (nsamples,nqoi+1) is the cost of the simulation. This column is not included in values.

cost_function(config_samples=None)[source]

Retrun the cost of evaluating the functions with the ids given in a set of config_samples. These samples are assumed to be in user space not canonical space

Parameters:
config_samplesnp.ndarray (nconfig_vars, nsamples)

The configuration indices