timescales.pipe.Pipe

class timescales.pipe.Pipe(n_seconds, fs, seeds=None)

Simulations, PSD/ACF, and fitting pipelines.

Parameters

n_secondsfloat

Length of signal, in seconds.

fsfloat

Sampling rate, in Hertz.

seeds1d array of int

Random seeds for reproducible simulations.

__init__(n_seconds, fs, seeds=None)

Methods

__init__(n_seconds, fs[, seeds])
add_step(step, *args, **kwargs)	Add a step to the pipeline.
bin(bin_size)	Bin signal.
fit(return_attrs, **fit_kwargs)	Fit timescale of simulation.
normalize(sig[, mean, variance])	Normalize signal with given mean and variance.
rescale(sig[, norm_range])	Normalize signal from lower to upper.
run([n_jobs, progress])	Run analysis pipeline.
sample([fs])	Sample binary array from probabilties.
simulate(sim_func, *sim_args[, operator, ...])	Simulate aperiodic signal.
transform(method, **compute_kwargs)	Fit timescale of simulation.

add_step(step, *args, **kwargs)

Add a step to the pipeline.

Parameters

step{‘simulate’, ‘sample’, ‘transform’, fit’}

Method to run.

*args

Positional arugments for the specified method.

**kwargs

Keyword arguemnts for the specified method.

bin(bin_size)

Bin signal.

Parameters

bin_sizeint

Number of samples per bin.

fit(return_attrs, **fit_kwargs)

Fit timescale of simulation.

Parameters

return_attrsstr or list of str or {‘knee_freq’, ‘tau’, ‘rsq’}

Model attributes to specifically store. These are attributes of PSD or ACF objects set upon fitting.

**fit_kwargs

Keyword arguments passed to the fit method of the PSD or ACF objects.

Notes

Assumes fit type based on transform method call.

static normalize(sig, mean=None, variance=None)

Normalize signal with given mean and variance.

static rescale(sig, norm_range=(0, 1))

Normalize signal from lower to upper.

run(n_jobs=- 1, progress=None)

Run analysis pipeline.

Parameters

n_jobsint

Number of jobs to run in parralel.

progress{None, ‘tqdm’, ‘tqdm.notebook’}

Specify whether to display a progress bar. Uses ‘tqdm’, if installed.

sample(fs=None)

Sample binary array from probabilties.

Parameters

fsint, optional, default: None

Updated sampling rate.

Notes

Assumes the sig attribute is the target probability array.

simulate(sim_func, *sim_args, operator='add', rescale=None, mean=None, variance=None, **sim_kwargs)

Simulate aperiodic signal.

Parameters

sim_funcfunc

Simulation function.

operator{‘add’, ‘mul’, ‘sub’, ‘div’} or {‘+’, ‘*’, ‘-’, ‘/’}

Operator to combine signals.

rescaletuple of (float, float), optional, default: None

Minimum and maximum y-values of simulation.

meanfloat, optional, default: None

Mean to normalize to.

variancefloat, opational, default: None

Variance to normalize to.

*sim_args

Additonal simulation positional arugments.

**sim_kwargs

Additional simulation keyword arguments.

Notes

Either rescale or (mean, std) should be passed for each simulation.

transform(method, **compute_kwargs)

Fit timescale of simulation.

Parameters

method{‘PSD’, ‘ACF’}

Fitting method.

**compute_kwargs

Additional arguments to pass to compute_spectrum() or compute_acf() methods in the PSD or ACF objects.