|
auspol
|
`#auspol` hash tagged retweet diffusions |
|
evently
|
Fitting Hawkes processes with AMPL |
|
fit_series()
|
Fit a Hawkes process or HawkesN process model on one or many event cascades
and learn model parameters. |
|
fits_dist_matrix()
|
Given a list of grouped fits, compute a distance matrix |
|
generate_features()
|
Given a list of group-fits produced by 'group_fit_series', this function generates features
for each group-fit by summarizing the fitted parameters. |
|
generate_series()
|
Main function to generate a Hawkes process sequence. It allows intermediary
saves and continuing a stopped simulation. Creates a CSV file with two
columns, each row is an event: (magnitude, time) |
|
generate_user_magnitude()
|
Function for sampling from the powerlaw distribution of user influence
it is the equivalent of the Richter-Gutenberg distribution in the Helmstetter model
the powerlaw distribution was determined from the twitter data, from the #retweets
alpha = 2.016, xmin = 1. Draw n values |
|
get_a1()
|
Calculating the expected size of first level of descendants |
|
get_branching_factor()
|
Branching factor is the expected number of events generated
by a single event. |
|
get_hawkes_neg_likelihood_value()
|
Compute the negative log-likelihood values of a given model on a list of given
event cascades. |
|
get_model_intensity_at()
|
Compute the intensity value of a given model at time t |
|
get_viral_score()
|
Viral score is the total reaction of the system to a single promotion,
i.e. the expected cascade size started by a single event of magnitude |
|
group_fit_series()
|
Given a list of cascades, this function fits each cascade individually by calling [fit_series].
If the given cascades are in a named list, the names will be regarded as groups and the result will be reformatted as a list of
group fits. |
|
new_hawkes()
|
Create a new hawkes model with given arguments |
|
parse_raw_tweets_to_cascades()
|
This function extracts cascades from a given jsonl file where each line is a tweet
json object. Please refer to the Twitter developer documentation:
https://developer.twitter.com/en/docs/tweets/data-dictionary/overview/tweet-object |
|
plot_event_series()
|
Plot a Hawkes process and its intensity function |
|
plot_kernel_function()
|
Plot the kernel functions of Hawkes processes |
|
predict_final_popularity()
|
Predict the final popularity (event count) of give histories and
its model parameters. |
|
prepare_tmp_file()
|
Prepare the temporary auxilixry files for AMPL |
|
set_tmp_folder()
|
Set up the folder for placing temporary files, defaults to /tmp |
|
setup_ampl()
|
Set up the AMPL environment by downloading an AMPL demo version and the compiled
ipopt binary. Only supports UNIX compatible OSs. |