=======
yannopt
=======
Optimization algorithms. Lots of them. This is a work in progress -- use at
your own risk of numerical underflow.
========
Examples
========
>>> from yannopt.optimizers import LBFGS
>>> from yannopt.learning_rates import BacktrackingLineSearch
>>> from yannopt.stopping_criteria import MaxIterations
>>> from yannopt.testing import check_optimizer
>>> from yannopt.testing import problems
>>>
>>>
>>> class Optimizer(BacktrackingLineSearch, MaxIterations, LBFGS):
>>> def __init__(self, n_iter):
>>> BacktrackingLineSearch.__init__(self, a=0.5)
>>> MaxIterations.__init__(self, n_iter)
>>> LBFGS.__init__(self, k=2)
>>>
>>> # Define objective function
>>> A = np.array([[1.0, 0.5, 0.0],
>>> [0.5, 1.0, 0.5],
>>> [0.0, 0.5, 1.0]])
>>> b = np.array([1.0, 2.0, 3.0])
>>> objective = f.Quadratic(A, b)
>>>
>>> # Define problem (objective + constraints)
>>> problem = minimize(objective)
>>> x0 = np.zeros(len(b))
>>>
>>> # Optimize!
>>> optimizer = Optimizer(20)
>>> solution = optimizer.optimize(problem, x0)
>>> print 'The minimizer of f(x) = 0.5 x'Ax - b is %s'.format(solution.x)