cbp.utils

cbp.utils.event_utils

cbp.utils.event_utils.compare_marginals(mar_1, mar_2)

cbp.utils.event_utils.diff_max_marginals(mar_1, mar_2)

cbp.utils.event_utils.engine_loop(engine_fun, max_iter=5000000, tolerance=0.01, error_fun=None, meassure_fun=None, isoutput=False, silent=False)

work engine for loop

Parameters

• engine_fun – work function

• max_iter (int, optional) – max iteration, defaults to 5000000

• tolerance (float, optional) – less than tolerance, stop, defaults to 1e-2

• error_fun (func, optional) – output the metrics for this iteration, defaults to None

• meassure_fun (func, optional) – relative distance between continuous two metrics, defaults to None

• isoutput (bool, optional) – output the distance, defaults to False

• silent (bool, optional) – output details, defaults to False

Returns

list of relative distance, num of running iteration, timers

cbp.utils.message

class cbp.utils.message.Message(sender, val)

Bases: cbp.utils.message.Message

cbp.utils.np_utils

cbp.utils.np_utils.batch_normal_angle(angle)

cbp.utils.np_utils.empirical_marginal(traj, num_bins)

cbp.utils.np_utils.nd_expand(inputdata, target_shape, expand_dim)

expand ndarray to target shape

Parameters

• inputdata (list or 1d ndarray) –

• target_shape (tuple) –

• expand_dim (int) – [description]

Returns

expanded ndarray with target_shape

Return type

ndarray

>>> inputdata = [1,2]
>>> target_shape = (1,2,3)
>>> expand_dim = 1
>>> output = np.array([
            [1,1,1],
            [2,2,2]
        ])

cbp.utils.np_utils.nd_multiexpand(input_data, target_shape, which_dims)

cbp.utils.np_utils.reduction_ndarray(ndarray, reduction_index)

reduct ndarray according to one index

Parameters

• ndarray (ndarray) – [description]

• reduction_index (int) – [description]

Returns

[description]

Return type

ndarray