class ImpliedTimescales

class deeptime.util.validation.ImpliedTimescales(lagtimes, its, its_stats=None)

Instances of this class hold a sequence of lagtimes and corresponding process timescales (potentially with process timescales of sampled models in a Bayesian setting). Objects can be used with plot_implied_timescales.

In case models over a range of lagtimes are available, the static method from_models() can be used.

Parameters:

• lagtimes (iterable of int) – Lagtimes corresponding to processes and their timescales.

• its (iterable of ndarray) – The timescales for each process, shape (n_lagtimes, n_processes).

• its_stats (list of list of ndarray, optional, default=None) – Sampled timescales of shape (n_lagtimes, n_processes, n_samples).

See also

deeptime.plots.plot_implied_timescales

Attributes

has_samples	Whether the data contains samples.
lagtimes	Yields the lagtimes corresponding to an instance of this class.
max_n_processes	Maximum number of processes.
max_n_samples	Maximum number of samples.
n_lagtimes	Number of lagtimes.

Methods

n_samples(lagtime_index, process_index)	Yields the number of samples for a particular lagtime and a particular process.
plot(*args, **kw)	Dispatches to plot_implied_timescales().
samples_for_process(process_index)	Yields timescales samples for a particular process.
timescales_for_process(process_index)	Yields maximum-likelihood timescales for a particular process.

n_samples(lagtime_index: int, process_index: int) → int

Yields the number of samples for a particular lagtime and a particular process.

Parameters:

• lagtime_index (int) – The lagtime index corresponding to lagtimes.

• process_index (int) – The process index.

Returns:

n_samples – The number of samples.

Return type:

int

plot(*args, **kw)

Dispatches to plot_implied_timescales().

samples_for_process(process_index: int) → ndarray

Yields timescales samples for a particular process.

Parameters:

process_index (int) – The process.

Returns:

timescales_samples – The sampled timescales for a particular process. Might contain NaN.

Return type:

ndarray(lagtimes, max_n_samples)

timescales_for_process(process_index: int) → ndarray

Yields maximum-likelihood timescales for a particular process.

Parameters:

process_index (int) – The process.

Returns:

timescales – The timescales for the particular process. Might contain NaN.

Return type:

ndarray (lagtimes,)

property has_samples: bool

Whether the data contains samples.

property lagtimes: ndarray

Yields the lagtimes corresponding to an instance of this class.

property max_n_processes: int

Maximum number of processes.

property max_n_samples: int

Maximum number of samples.

property n_lagtimes: int

Number of lagtimes.