ivneuro.events

User-facing functions for processing events.

Functions

`make_intervals`(timestamps, start, end)	Make intervals based on timestamps.
`classify_events_base_on_time`(event1, event2, treshold)	Classify an event in two categories based on how close in time it occurs from an event of reference.

Module Contents

ivneuro.events.make_intervals(timestamps, start, end)

Make intervals based on timestamps.

Parameters:

timestamps (np.ndarray or list) – Timestamps to use as reference to make intervals.
start (int or float) – Start of interval, relative to timestamps.
end (int or float) – End of interval, relative to timestamps.

Returns:

Slices with the start time and end time of each interval.

Return type:

list

Examples

Create intervals using make_intervals function.

>>> import ivneuro as ivn
>>> event = [*range(10,40, 10)] # Timestamps
>>> # Make intervals
>>> intervals = ivn.make_intervals(event, 0, 3)
>>> print(intervals)
[slice(10, 13, None), slice(20, 23, None), slice(30, 33, None)]

Use intervals to slice a pandas.DataFrame

>>> # Create dataframe
>>> import numpy as np
>>> import pandas as pd
>>> np.random.seed(24)
>>> df=pd.DataFrame(data=np.random.rand(40), columns=['values'])
>>> # Slice datafrane
>>> [df.loc[i,] for i in intervals]
[      values
 10  0.320519
 11  0.366415
 12  0.709652
 13  0.900142,
       values
 20  0.842780
 21  0.306013
 22  0.631170
 23  0.680239,
       values
 30  0.486032
 31  0.807219
 32  0.844220
 33  0.534681]

ivneuro.events.classify_events_base_on_time(event1, event2, treshold, mode='left')

Classify an event in two categories based on how close in time it occurs from an event of reference.

Parameters:

event1 (numpy.array of shape (1 x n)) – Event to classify.
event2 (numpy.array of shape (1 x m)) – Event of reference.
treshold (TYPE) – Threshold amount of time used to classify events.
mode (str, optional) – Define the mode of evaluation of proximity. “left”, only looks event1 that occur before event2; “right”, only looks event1 that ocurr after event2; “two-sides”, look temporal proximity before and after. The default is ‘left’.

Returns:

near (np.array of shape (1 x o)) – Subset of event1 classified as temporally close to event2.
far (np.array of shape (1 x p).) – Subset of event1 classified as temporally far from event2.

Examples

Classify events with classify_events_base_on_time function.

>>> # Import packages and create events
>>> import ivneuro as ivn
>>> import numpy as np
>>> event1 = np.array([3, 7, 10.5, 15.3])
>>> event2 = np.array([1, 7.5, 15])
>>> # Classify events
>>> near, far = ivn.classify_events_base_on_time(event1, event2, treshold = 1)
>>> print(near)
[7.]
>>> print(far)
[ 3.  10.5 15.3]

Set mode to ‘right’.

>>> near, far = ivn.classify_events_base_on_time(event1, event2, treshold = 1, mode = 'right')
>>> print(near)
[15.3]
>>> print(far)
[ 3.   7.  10.5]

Set mode to ‘both’.

>>> near, far = ivn.classify_events_base_on_time(event1, event2, treshold = 1, mode = 'two-sides')
>>> print(near)
[ 7.  15.3]
>>> print(far)
[ 3.  10.5]