ddf.measurement package

Submodules

ddf.measurement.gradient_storage module

class ddf.measurement.gradient_storage.GradientStorage(storage: dict[ddf.ant_link.loss_channel.LossChannel, numpy.ndarray])

Bases: object

storage: dict[LossChannel, ndarray]

ddf.measurement.measurement module

class ddf.measurement.measurement.Measurement(parent_object_id: ddf.ddf.ObjectId, operator: ddf.operator.Operator, exchange_vector: ddf.information.ExchangeVector, state_jacobian: numpy.ndarray, source: ddf.ant_link.ant_link.AntLink, gradient_storage: ddf.measurement.gradient_storage.GradientStorage = <factory>)

Bases: object

backpropagate(node: Node, loss_gradient_wrt_state: ndarray, loss_channel: LossChannel) → None

exchange_vector: ExchangeVector

contains b and cov_b

get_state_info_mat(state: InternalState) → ndarray

Returns information matrix in state space (not exchange state space).

gradient_storage: GradientStorage

operator: Operator

optimize_step() → None

parent_object_id: ObjectId

object id of the parent AntObjectReference. This could be replaced with a direct pointer to the parent.

set_gradients_to_zero() → None

source: AntLink

state_jacobian: ndarray

Jacobian of the AntObjectReference state w.r.t. the exchange_state of this measurement.

Module contents