#!/usr/bin/env python
"""Copyright (C) 2012 Computational Neuroscience Group, NMBU.
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
"""
from __future__ import division
import numpy as np
import scipy.stats
import warnings
[docs]def get_activation_times_from_distribution(n, tstart=0., tstop=1.E6,
distribution=scipy.stats.expon,
rvs_args=dict(loc=0, scale=1),
maxiter=1E6):
"""
Construct a length n list of ndarrays containing continously increasing
random numbers on the interval [tstart, tstop], with intervals drawn from
a chosen continuous random variable distribution subclassed from
scipy.stats.rv_continous, e.g., scipy.stats.expon or scipy.stats.gamma.
The most likely initial first entry is
``tstart + method.rvs(size=inf, **rvs_args).mean()``
Parameters
----------
n : int
number of ndarrays in list
tstart : float
minimum allowed value in ndarrays
tstop : float
maximum allowed value in ndarrays
distribution : object
subclass of scipy.stats.rv_continous. Distributions
producing negative values should be avoided if continously increasing
values should be obtained, i.e., the probability density function
``(distribution.pdf(**rvs_args))`` should be ``0`` for ``x < 0``, but this is not
explicitly tested for.
rvs_args : dict
parameters for method.rvs method. If "size" is in dict, then tstop will
be ignored, and each ndarray in output list will be
``distribution.rvs(**rvs_args).cumsum() + tstart``. If size is not given in dict,
then values up to tstop will be included
maxiter : int
maximum number of iterations
Returns
-------
list of ndarrays
length n list of arrays containing data
Raises
------
AssertionError
if distribution does not have the 'rvs' attribute
StopIteration
if number of while-loop iterations reaches maxiter
Examples
--------
Create n sets of activation times with intervals drawn from the exponential
distribution, with rate expectation lambda 10 s^-1 (thus
scale=1000 / lambda). Here we assume output in units of ms
>>> from LFPy.inputgenerators import get_activation_times_from_distribution
>>> import scipy.stats as st
>>> import matplotlib.pyplot as plt
>>> times = get_activation_times_from_distribution(n=10, tstart=0., tstop=1000.,
>>> distribution=st.expon,
>>> rvs_args=dict(loc=0.,
>>> scale=100.))
"""
try:
assert hasattr(distribution, 'rvs')
except AssertionError:
raise AssertionError('distribution={} must have the attribute "rvs"'.format(distribution))
times = []
if 'size' in rvs_args.keys():
for i in range(n):
times += [distribution.rvs(**rvs_args).cumsum() + tstart]
else:
for i in range(n):
values = distribution.rvs(size=1000, **rvs_args).cumsum() + tstart
iter = 0
while values[-1] < tstop and iter < maxiter:
values = np.r_[values, distribution.rvs(size=1000, **rvs_args).cumsum() + values[-1]]
iter += 1
if iter == maxiter:
raise StopIteration('maximum number of iterations reach. Con')
times += [values[values < tstop]]
return times