"""Common function to run online and offline algorithms."""
import time
import numpy as np
[docs]def run_algorithm(opt, max_nb_of_iter, verbose=0):
"""Run the algorithm setup with the defined optimizer.
Parameters
----------
opt: Optimizer Class
max_nb_of_iter: int
Maximum number of iteration
verbose: int
Verbosity level.
Returns
-------
x_final: numpy.ndarray
the estimated solution.
costs: list of float
the cost function values.
metrics: dict
the requested metrics values during the optimization.
"""
start = time.perf_counter()
cost_op = opt._cost_func
# Perform the reconstruction
if verbose > 0:
print("Starting optimization...")
opt.iterate(max_iter=max_nb_of_iter)
end = time.perf_counter()
if verbose > 0:
if hasattr(cost_op, "cost"):
print(" - final iteration number: ", cost_op._iteration)
print(" - final log10 cost value: ", np.log10(cost_op.cost))
print(" - converged: ", opt.converge)
print("Done.")
print("Execution time: ", end - start, " seconds")
print("-" * 40)
if hasattr(opt._grad, "linear_op"):
x_final = opt._grad.linear_op.adj_op(opt.x_final)
else:
x_final = opt.x_final
metrics = opt.metrics
if hasattr(cost_op, "cost"):
costs = cost_op._cost_list
else:
costs = None
return x_final, costs, metrics
[docs]def run_online_algorithm(opt, kspace_generator, estimate_call_period=None, verbose=0):
"""Run online optimisation algorithm.
At each step the obs_data is updated via the kspace_generator.
Parameters
----------
opt: instance of SetUp
optimisation algorithm instance
kspace_generator: instance of BaseKspaceGenerator
The kspace_generator yielding the observed data to be updated.
estimate_call_period: int, default None
The period over which to retrieve an estimate of the online algorithm.
If None, only the last estimate is retrieved.
"""
opt.idx = 0
cost_op = opt._cost_func
# Perform the first reconstruction
if verbose > 0:
print("Starting optimization...")
start = time.perf_counter()
estimates = list()
estimates += kspace_generator.opt_iterate(opt, estimate_call_period=estimate_call_period)
end = time.perf_counter()
if verbose > 0:
if hasattr(cost_op, "cost"):
print(" - final iteration number: ", cost_op._iteration)
print(" - final cost value: ", cost_op.cost)
print(" - converged: ", opt.converge)
print("Done.")
print("Execution time: ", end - start, " seconds")
print("-" * 40)
# Get the final solution
if hasattr(opt._grad, "linear_op"):
x_final = opt._grad.linear_op.adj_op(opt._x_new)
else:
x_final = opt._x_new
costs = cost_op._cost_list
metrics = {}
metrics.update(opt.metrics)
metrics.update({'estimates': estimates})
return x_final, costs, metrics
