dadra package

Subpackages

• dadra.systems package
  • dadra.systems.dynamics module
  • dadra.systems.samplers module
• dadra.utils package
  • dadra.utils.christoffel_utils module
  • dadra.utils.graph_utils module
  • dadra.utils.misc_utils module
  • dadra.utils.p_utils module

dadra.disturbance module

dadra.dyn_sys module

class dadra.dyn_sys.DisturbedSystem(dyn_func, intervals, state_dim, disturbance: dadra.disturbance.Disturbance, timesteps=100, parts=1001, all_time=False)

Bases: dadra.dyn_sys.System

sample_system(N)

Draws N samples from the specified system

Parameters

N (int) – The number of samples to be drawn

Returns

Array of N samples

Return type

numpy.ndarray

system_sampler(x=None)

Obtains sample from specified system with disturbance

Parameters

x (NoneType, optional) – Placeholder variable over which to evaluate, defaults to None

Returns

The sample from the specified system at the last timestep if the all_time is False and including all timesteps otherwise

class dadra.dyn_sys.Sampler(sample_fn, state_dim, timesteps, parts, all_time)

Bases: dadra.dyn_sys.System

class implementation of a dynamical system with a means of sampling the system specified by the user

Parameters

• sample_fn (function) – A function to sample from

• state_dim (int) – The degrees of freedom of the system

• timesteps (int, optional) – The number of timesteps over which to compute the sample, defaults to 100

• parts (int, optional) – The number of parts to partition the time interval into for computing the sample, defaults to 1001

• all_time (bool, optional) – If True, each sample will include all timesteps from the system, rather than only the last timestep, defaults to False

sample_system(N)

Draws N samples from the specified system

Parameters

N (int) – The number of samples to be drawn

Returns

Array of N samples

Return type

numpy.ndarray

class dadra.dyn_sys.SimpleSystem(dyn_func, intervals, state_dim, timesteps=100, parts=1001, all_time=False)

Bases: dadra.dyn_sys.System

Class implementation of a dynamical system that allows for parallelized sampling.

Parameters

• dyn_func (function) – The function defining the dynamics of the system

• intervals (list) – The intervals corresponding to the possible values of the initial states of the variables in the system

• state_dim (int) – The degrees of freedom of the system

• timesteps (int, optional) – The number of timesteps over which to compute the sample, defaults to 100

• parts (int, optional) – The number of parts to partition the time interval into for computing the sample, defaults to 1001

• all_time (bool, optional) – If True, each sample will include all timesteps from the system, rather than only the last timestep, defaults to False

static dyn_to_sys(arr, t, func_list, *args)

Static method which allows for defining the dynamics of a system from an initial state, the time, and a list of functions for each dimension

Parameters

• arr (numpy.ndarray) – Array input, corresponding to the initial state

• t (float) – The time

• func_list (list) – A list of functions, one for each dimension

Returns

The functional dynamics of the system

Return type

list

classmethod from_list(func_list, intervals, timesteps=100, parts=1001, all_time=False)

Class method that allows for an instance of dadra.DynamicSystem to be initialized using a list of functions, one for each variable, to define the dynamics of a system rather than a single function

Parameters

• func_list (list) – The list of functions, one for each variable, that define the dynamics of the system

• intervals (list) – The intervals corresponding to the possible values of the initial states of the variables in the system

• timesteps (int, optional) – The number of timesteps over which to compute the sample, defaults to 100

• parts (int, optional) – The number of parts to partition the time interval into for computing the sample, defaults to 1001

• all_time (bool, optional) – If True, each sample will include all timesteps from the system, rather than only the last timestep, defaults to False

Raises

• TypeError – If not all functions in func_list are callable

• ValueError – If not all functions in func_list include parameters for the intial state and time

• ValueError – If not all functions in func_list include parameters for the intial state, time, and the provided additional arguments

Returns

A dadra.DynamicSystem object

Return type

dadra.DynamicSystem

sample_system(N)

Draws N samples from the specified system

Parameters

N (int) – The number of samples to be drawn

Returns

Array of N samples

Return type

numpy.ndarray

system_sampler(x=None)

Obtains sample from specified system

Parameters

x (NoneType, optional) – Placeholder variable over which to evaluate, defaults to None

Returns

The sample from the specified system at the last timestep if the all_time is False and including all timesteps otherwise

Return type

numpy.ndarray

class dadra.dyn_sys.System(dyn_func, state_dim, intervals)

Bases: object

Parent class that serves as an interface for dynamic systems objects

Parameters

• dyn_func (function) – The function defining the dynamics of the system

• state_dim (int) – The degrees of freedom of the system

• intervals (list) – The intervals corresponding to the possible values of the initial states of the variables in the system

check_dyn_func()

Checks whether the dynamic function in this instance of the dadra.DynamicSystem class is valid

Raises

• TypeError – If the provided dynamic function is not callable

• ValueError – If there is an invalid number of required arguments in the dynamic function

check_intervals()

Checks whether the list of intervals in this instance of the dadra.DynamicSystem class is valid

Raises

• ValueError – If there is not an interval defined for each variable

• ValueError – If not all intervals include an upper and a lower bound

• ValueError – If not all intervals consist of two numbers

sample_system(N)

Draws N samples from the specified system

Parameters

N (int) – The number of samples to be drawn

Returns

Array of N samples

Return type

numpy.ndarray

system_sampler(x=None)

Obtains sample from specified system

Parameters

x (NoneType, optional) – Placeholder variable over which to evaluate, defaults to None

Returns

The sample from the specified system at the last timestep

Return type

numpy.ndarray

dadra.estimate module

dadra.sampling module

dadra.sampling.make_sample_n(sample_fn, parallel=True, pool=None)

Takes in a sample function and allows for parallelized sampling

Parameters

• sample_fn (function) – A function to compute samples

• parallel (bool, optional) – True if parallelization is to be used to compute samples, defaults to True

• pool (multiprocessing.pool.Pool, optional) – Pool to use for parallelization if specified, defaults to None