echotorch.datasets package

Submodules

echotorch.datasets.MackeyGlassDataset module

class echotorch.datasets.MackeyGlassDataset.MackeyGlassDataset(sample_len, n_samples, tau=17, seed=None)

Bases: torch.utils.data.dataset.Dataset

Mackey Glass dataset

echotorch.datasets.MemTestDataset module

class echotorch.datasets.MemTestDataset.MemTestDataset(sample_len, n_samples, n_delays=10, seed=None)

Bases: torch.utils.data.dataset.Dataset

Generates a series of input timeseries and delayed versions as outputs. Delay is given in number of timesteps. Can be used to empirically measure the memory capacity of a system.

echotorch.datasets.NARMADataset module

class echotorch.datasets.NARMADataset.NARMADataset(sample_len, n_samples, system_order=10, seed=None)

Bases: torch.utils.data.dataset.Dataset

xth order NARMA task WARNING: this is an unstable dataset. There is a small chance the system becomes unstable, leading to an unusable dataset. It is better to use NARMA30 which where this problem happens less often.

Module contents

class echotorch.datasets.DatasetComposer(datasets, *args, **kwargs)

Bases: torch.utils.data.dataset.Dataset

Compose dataset

class echotorch.datasets.HenonAttractor(sample_len, n_samples, xy, a, b, washout=0, normalize=False, seed=None)

Bases: torch.utils.data.dataset.Dataset

The Rössler attractor is the attractor for the Rössler system, a system of three non-linear ordinary differential equations originally studied by Otto Rössler. These differential equations define a continuous-time dynamical system that exhibits chaotic dynamics associated with the fractal properties of the attractor.

regenerate()

Regenerate :return:

class echotorch.datasets.LambdaDataset(sample_len, n_samples, func, start=0, dtype=torch.float32)

Bases: torch.utils.data.dataset.Dataset

Create simple periodic signal timeseries

class echotorch.datasets.LogisticMapDataset(sample_len, n_samples, alpha=5, beta=11, gamma=13, c=3.6, b=0.13, seed=None)

Bases: torch.utils.data.dataset.Dataset

Logistic Map dataset

class echotorch.datasets.LorenzAttractor(sample_len, n_samples, xyz, sigma, b, r, dt=0.01, washout=0, normalize=False, seed=None)

Bases: torch.utils.data.dataset.Dataset

The Rössler attractor is the attractor for the Rössler system, a system of three non-linear ordinary differential equations originally studied by Otto Rössler. These differential equations define a continuous-time dynamical system that exhibits chaotic dynamics associated with the fractal properties of the attractor.

regenerate()

Regenerate :return:

class echotorch.datasets.MackeyGlassDataset(sample_len, n_samples, tau=17, seed=None)

Bases: torch.utils.data.dataset.Dataset

Mackey Glass dataset

class echotorch.datasets.MemTestDataset(sample_len, n_samples, n_delays=10, seed=None)

Bases: torch.utils.data.dataset.Dataset

Generates a series of input timeseries and delayed versions as outputs. Delay is given in number of timesteps. Can be used to empirically measure the memory capacity of a system.

class echotorch.datasets.NARMADataset(sample_len, n_samples, system_order=10, seed=None)

Bases: torch.utils.data.dataset.Dataset

xth order NARMA task WARNING: this is an unstable dataset. There is a small chance the system becomes unstable, leading to an unusable dataset. It is better to use NARMA30 which where this problem happens less often.

class echotorch.datasets.RosslerAttractor(sample_len, n_samples, xyz, a, b, c, dt=0.01, washout=0, normalize=False, seed=None)

Bases: torch.utils.data.dataset.Dataset

The Rössler attractor is the attractor for the Rössler system, a system of three non-linear ordinary differential equations originally studied by Otto Rössler. These differential equations define a continuous-time dynamical system that exhibits chaotic dynamics associated with the fractal properties of the attractor.

regenerate()

Regenerate :return:

class echotorch.datasets.SinusoidalTimeseries(sample_len, n_samples, period, a=1.0, m=0.0, start=1, dtype=torch.float64)

Bases: torch.utils.data.dataset.Dataset

Sinusoidal timeseries

random_initial_points()

Random initial points :return:

regenerate()

Regenerate :return:

class echotorch.datasets.PeriodicSignalDataset(sample_len, n_samples, period, start=1, dtype=torch.float64)

Bases: torch.utils.data.dataset.Dataset

Create simple periodic signal timeseries