Event Management

This modules provides methods for event handling.

Every event is an dictionary in the following form:

event = {
    "time": [a float value],
    "sender": [a single sender (int or str) or a list of senders],
    "receiver": [a single receiver (int or str) or a list of receivers] (optional),
}

Tools

This file provides tools for event handling.

class dynamix.events.tools.MakeID

Converter object for sender/receiver labels.

Methods

__call__(generator)	Replaces all sender and receiver values by integer keys.
make_id(generator)	Replaces all sender and receiver values by integer keys.

make_id(generator)

Replaces all sender and receiver values by integer keys.

Parameters:

generator : iterable or single event

Events

dynamix.events.tools.event_to_string(event)

Converts the Event-Dictionary into a string representation.

Parameters:

event : dict

Event in dict representation

Returns:

line : str

String representation of the event

Notes

Example:

>>> event_to_string({"time":1,"sender":[1,2,3],"receiver":[4,5,6]})
'1;1,2,3;4,5,6'

>>> event_to_string({"time":1,"sender":[1,2,3]})
'1;1,2,3'

>>> event_to_string({"time":1,"sender":1,"receiver":2})
'1;1;2'

dynamix.events.tools.expand_sender_events(generator, directed=True)

Expands sender events from each sender to each other sender

The functions uses an iterator of events and creates for every sender-sender combination a new event.

Parameters:

generator : iterable or single event

Events

directed : boolean

If true generates A->B and B->A

dynamix.events.tools.jitter_equal(generator, interval=(-1, 1))

Adds noise to the event time.

This functions adds an equal distributed noise in the given interval to the time of the events.

Parameters:

generator : iterable or single event

Events

interval : tuple or list

2 element tuple with min and max jitter

Notes

Events could be out of order after adding jitter.

dynamix.events.tools.load(filename)

Loads events from file

Parameters:

filename : str

Filename which is used to load the data.

dynamix.events.tools.make_lists(event)

Converts all sender and receiver literals into lists

dynamix.events.tools.make_literals(event)

Converts all sender and receiver lists into literals if they contain only one element

dynamix.events.tools.random() → x in the interval [0, 1).

dynamix.events.tools.save(generator, filename)

Saves events to file

Parameters:

generator : iterable or single event

Events

filename : str

Filename which is used to save the file.

dynamix.events.tools.simplify(generator)

Simplifies the compact representation with multiple sender and receivers.

The functions uses an iterator of events and creates for every sender-receiver combination a new event.

Parameters:

generator : iterable or single event

Events

dynamix.events.tools.sort(generator, window_size=5.0)

Sorts events in an event stream.

Elements of an generator will be sorted in the given time window size.

Parameters:

generator : iterable or single event

Events

window_size : int or float

time window size

Notes

Events would be delayed until next event with timedelta > window_size arrives.

dynamix.events.tools.string_to_event(string)

Converts the string representation of an event into the dictionary representation.

Parameters:

string : str

Event in string representation

Returns:

event : dict

Dictionary representation of the event

Notes

Example:

>>> event = string_to_event('1;1,2,3;4,5,6')
>>> event ==  {'time': 1, 'sender': [1, 2, 3], 'receiver': [4, 5, 6]}
True

>>> event = string_to_event('1;1,2,3')
>>> event == {'time': 1, 'sender': [1, 2, 3]}
True

>>> event = string_to_event('1;1;2')
>>> event == {'time': 1, 'sender': 1, 'receiver': 2}
True

dynamix.events.tools.throttle(generator, factor=1.0)

Throttles event processing

The events will be repressed until the passing time from the first event to the current one is at least factor times the event time between the first event and the current one.

Parameters:

generator : iterable or single event

Events

factor : float

factor of time delay, default 1.0

Notes

The time will be interpreted as seconds.

Twitter - Stream

Creates Events from Twitter.com

class dynamix.events.twitter.TwitterEvents(auth)

Twitter api wrapper

Attributes

auth

Methods

hashtags([limit, min_tags, search_tags])	Generates events from Twitter Hashtags.
mentions([limit])	Generates events from Twitter mentions.
raw_tweets()	Iterator for basic tweet stream.
user_activities([limit])	Generates events from Twitter Tweets.

hashtags(limit=None, min_tags=2, search_tags=None)

Generates events from Twitter Hashtags.

Every tweet with hashtags generates an event with sender is a list of all containing hashtags and receiver not set.

Parameters:

limit : int or None

Limits the result size. Default is None - no limit.

min_tags : int

Minimum number of hashtags (excluding search tags)

search_tags : list of strings or string

List of hashtags. At least one hashtag must be in the tweet

mentions(limit=None)

Generates events from Twitter mentions.

Every tweet with mentions generates an event with sender is the creator of the tweed and receiver is a list of mentions.

Parameters:

limit : int or None

Limits the result size. Default is None - no limit.

raw_tweets()

Iterator for basic tweet stream.

user_activities(limit=None)

Generates events from Twitter Tweets.

Every tweet generates an event with sender is the creator of the tweet.

Parameters:

limit : int or None

Limits the result size. Default is None - no limit.

dynamix.events.twitter.get_tweet_timestamp(tweet)

Extract the unix timestamp from Tweet

Parameters:

tweet : dict

Twitter tweet in dict format