****
## Explanation on the project


This is the code used for the paper Why Rumors Spread Fast in Social Networks, and How to Stop It

## Dependencies and Data
To run the code, please make sure to download the datasets
from the following sites:

Facebook: http://snap.stanford.edu/data/ego-Facebook.html \
Twitter: http://snap.stanford.edu/data/ego-Twitter.html \
Google+: http://snap.stanford.edu/data/ego-Gplus.html \
Twitch: http://snap.stanford.edu/data/twitch-social-networks.html

The local paths are set in the init of the module IMP

To use this code, one should install the dependencies that you find
listed in the file _requirements.txt._ file in the code directory.

It would be easy if these would be installed in a virtual environment

#install pipenv if you don't have this
pip install --user pipenv
pipenv install

To update the requirements file:
pipenv lock -r > requirements.txt

However, if you don't want to set up a virtual environment you can
still download the requirements as follows
#make sure to run this command in the code directory
pip install -r requirements.txt


## Code Structure
The module IMP (which contains all the code) consists of four files
- main: from which the code is run
- simulation: the implementation of our rumor spreading model and all the counter measures
- averaging: a function that generates the correct graph structure based on the social network considered and a dummy
function that repeats the experiments several times in the case that it involves randomness and takes the average of all
experiments.

dict_counter_measure = {'id':COUNTER_MEASURE_FACTCHECKING, 'num_green':1, 'start_time':4, 'green_spreading_ratio':0.5,
                            'p_orange_h':0.05, 'p_orange_r':0.1, 'p_coinflip':0.7, 'frac_green':0.1, 'threshold_detection':0.05,
                            'threshold_block':0.05}
    dict_args = {"degree_of_supernodes":150, "number_of_supernodes": 1000, "nodes_in_clique": 16}

## How to run the experiments
- To obtain the results for the Flower graph and no counter measures:
Run main with parameters
num_runs=50,
graph_loc=fb_loc (this will be overwritten we do not use this)
type_graph="FlowerGraph",
num_red=1,
k=5,
dict_args={"num_cliques": 1000, "nodes_in_clique": 16}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}


- ME graph and no countermeasures for D=50 (used in Figure 1a)
num_runs=50,
graph_loc=fb_loc (this will be overwritten we do not use this)
type_graph='moderatelyExpander',
num_red=1,
k=5,
dict_args={"degree_of_supernodes":50, "number_of_supernodes": 1000, "nodes_in_clique": 16}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

- ME graph and no countermeasure for D=150
num_runs=50,
graph_loc=fb_loc (this will be overwritten we do not use this)
type_graph='moderatelyExpander',
num_red=1,
k=5,
dict_args={"degree_of_supernodes":150, "number_of_supernodes": 1000, "nodes_in_clique": 16}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

- ER-high (used in Figure 1a) and no countermeasure
num_runs=50,
graph_loc=fb_loc (this will be overwritten we do not use this)
type_graph='ER',
num_red=1,
k=5,
dict_args={"num_nodes":16000, "p":4/ math.sqrt{16000}}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

- ER-low (used in Figure 1a) and no countermeasure
num_runs=50,
graph_loc=fb_loc (this will be overwritten we do not use this)
type_graph='ER',
num_red=1,
k=5,
dict_args={"num_nodes":16000, "p":1/(4 * math.sqrt{16000})}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

- FB SN no countermeasure
num_runs=50,
graph_loc=fb_loc
type_graph='SN',
num_red=1,
k=5,
dict_args={}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

- FB HRG no countermeasure
num_runs=50,
graph_loc=fb_loc
type_graph='HRG',
num_red=1,
k=5,
dict_args={}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

TW SN no countermeasure
num_runs=50,
graph_loc=twitter_loc
type_graph='SN',
num_red=1,
k=5,
dict_args={}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

TW HRG no countermeasure
num_runs=50,
graph_loc=twitter_loc
type_graph='HRG',
num_red=1,
k=5,
dict_args={}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

T-GE SN no countermeasure
num_runs=50,
graph_loc=musea_DE_loc
type_graph='SN',
num_red=1,
k=5,
dict_args={}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

T-GE HRG no countermeasure
num_runs=50,
graph_loc=musea_DE_loc
type_graph='HRG',
num_red=1,
k=5,
dict_args={}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

T-FR SN no countermeasure
num_runs=50,
graph_loc=musea_FR_loc
type_graph='SN',
num_red=1,
k=5,
dict_args={}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

T-FR HRG no countermeasure
num_runs=50,
graph_loc=musea_FR_loc
type_graph='SN',
num_red=1,
k=5,
dict_args={}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

G+ SN no countermeasure
num_runs=50,
graph_loc=Gplus_loc
type_graph='SN',
num_red=1,
k=5,
dict_args={}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

G+ HRG no countermeasure
num_runs=50,
graph_loc= Gplus_loc
type_graph='SN',
num_red=1,
k=5,
dict_args={}
dict_counter_measure={'id':COUNTER_MEASURE_NONE}

FB-CM1
num_runs=50,
graph_loc=fb_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ORANGE_NODES, 'p_orange_h':0.05,'p_orange_r':0.1}
dict_args = {}

TW-CM1
num_runs=50,
graph_loc=twitter_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ORANGE_NODES, 'p_orange_h':0.05,'p_orange_r':0.1}
dict_args = {}

ME-CM1
num_runs=50,
graph_loc=fb_loc (will be overwritten)
type_graph='moderatelyExpander',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ORANGE_NODES, 'p_orange_h':0.05,'p_orange_r':0.1}
dict_args = {"degree_of_supernodes":150, "number_of_supernodes": 1000, "nodes_in_clique": 16}

T-GE-CM1
num_runs=50,
graph_loc=musea_DE_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ORANGE_NODES, 'p_orange_h':0.05,'p_orange_r':0.1}
dict_args = {}

T-FR-CM1
num_runs=50,
graph_loc=musea_FR_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ORANGE_NODES, 'p_orange_h':0.05,'p_orange_r':0.1}
dict_args = {}

G+-CM1
num_runs=50,
graph_loc=Gplus_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ORANGE_NODES, 'p_orange_h':0.05,'p_orange_r':0.1}
dict_args = {}

FB-CM2
num_runs=10,
graph_loc=fb_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_COMMUNITY_DETECTION,'threshold_detection':0.05,
                            'threshold_block':0.05}
dict_args = {}

TW-CM2
num_runs=10,
graph_loc=twitter_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_COMMUNITY_DETECTION,'threshold_detection':0.05,
                            'threshold_block':0.05}
dict_args = {}

ME-CM2
num_runs=10,
graph_loc=fb_loc (will be overwritten)
type_graph='moderatelyExpander',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_COMMUNITY_DETECTION,'threshold_detection':0.05,
                            'threshold_block':0.05}
dict_args = {"degree_of_supernodes":150, "number_of_supernodes": 1000, "nodes_in_clique": 16}

T-GE-CM2
num_runs=10,
graph_loc= musea_DE_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_COMMUNITY_DETECTION,'threshold_detection':0.05,
                            'threshold_block':0.05}
dict_args = {}

T-FR-CM2
num_runs=10,
graph_loc= musea_FR_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_COMMUNITY_DETECTION,'threshold_detection':0.05,
                            'threshold_block':0.05}
dict_args = {}

G+-CM2
num_runs=10,
graph_loc= Gplus_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_COMMUNITY_DETECTION,'threshold_detection':0.05,
                            'threshold_block':0.05}
dict_args = {}

FB-CM3
num_runs=50,
graph_loc=fb_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ACCURACYFLAG, 'p_coinflip':0.7}
dict_args = {}

TW-CM3
num_runs=50,
graph_loc=twitter_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ACCURACYFLAG, 'p_coinflip':0.7}
dict_args = {}

ME-CM3
num_runs=50,
graph_loc=fb_loc (will be overwritten)
type_graph='moderatelyExpander',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ACCURACYFLAG, 'p_coinflip':0.7}
dict_args = {"degree_of_supernodes":150, "number_of_supernodes": 1000, "nodes_in_clique": 16}

T-GE-CM3
num_runs=50,
graph_loc=musea_DE_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ACCURACYFLAG, 'p_coinflip':0.7}
dict_args = {}

T-FR-CM3
num_runs=50,
graph_loc=musea_FR_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ACCURACYFLAG, 'p_coinflip':0.7}
dict_args = {}

G+-CM3
num_runs=50,
graph_loc=Gplus_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_ACCURACYFLAG, 'p_coinflip':0.7}
dict_args = {}

FB-CM4
num_runs=50,
graph_loc=fb_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_GREEN_NODES, 'num_green':1, 'start_time':4, 'green_spreading_ratio':0.5}
dict_args = {}

TW-CM4
num_runs=50,
graph_loc=twitter_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_GREEN_NODES, 'num_green':1, 'start_time':4, 'green_spreading_ratio':0.5}
dict_args = {}

ME-CM4
num_runs=50,
graph_loc=fb_loc (will be overwritten)
type_graph='moderatelyExpander',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_GREEN_NODES, 'num_green':1, 'start_time':4, 'green_spreading_ratio':0.5}
dict_args = {"degree_of_supernodes":150, "number_of_supernodes": 1000, "nodes_in_clique": 16}

T-GE-CM4
num_runs=50,
graph_loc=musea_DE_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_GREEN_NODES, 'num_green':1, 'start_time':4, 'green_spreading_ratio':0.5}
dict_args = {}

T-FR-CM4
num_runs=50,
graph_loc=musea_FR_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_GREEN_NODES, 'num_green':1, 'start_time':4, 'green_spreading_ratio':0.5}
dict_args = {}

G+-CM4
num_runs=50,
graph_loc=twitter_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_GREEN_NODES, 'num_green':1, 'start_time':4, 'green_spreading_ratio':0.5}
dict_args = {}

FB-CM5
num_runs=50,
graph_loc=fb_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_FACTCHECKING, 'frac_green':0.1}
dict_args = {}

TW-CM5
num_runs=50,
graph_loc=twitter_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_FACTCHECKING, 'frac_green':0.1}
dict_args = {}

ME-CM5
num_runs=50,
graph_loc=fb_loc (will be overwritten)
type_graph='moderatelyExpander',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_FACTCHECKING, 'frac_green':0.1}
dict_args = {"degree_of_supernodes":150, "number_of_supernodes": 1000, "nodes_in_clique": 16}

T-GE-CM5
num_runs=50,
graph_loc=musea_DE_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_FACTCHECKING, 'frac_green':0.1}
dict_args = {}

T-FR-CM5
num_runs=50,
graph_loc=musea_FR_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_FACTCHECKING, 'frac_green':0.1}
dict_args = {}

G+-CM5
num_runs=50,
graph_loc=Gplus_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_FACTCHECKING, 'frac_green':0.1}
dict_args = {}

FB-CM6
num_runs=50,
graph_loc=fb_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_HEAR_FROM_AT_LEAST_TWO}
dict_args = {}

TW-CM6
num_runs=50,
graph_loc=twitter_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_HEAR_FROM_AT_LEAST_TWO}
dict_args = {}

ME-CM6
num_runs=50,
graph_loc=fb_loc (will be overwritten)
type_graph='moderatelyExpander',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_HEAR_FROM_AT_LEAST_TWO}
dict_args = {"degree_of_supernodes":150, "number_of_supernodes": 1000, "nodes_in_clique": 16}


T-GE-CM6
num_runs=50,
graph_loc=musea_DE_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_HEAR_FROM_AT_LEAST_TWO}
dict_args = {}


T-FR-CM6
num_runs=50,
graph_loc=musea_FR_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_HEAR_FROM_AT_LEAST_TWO}
dict_args = {}


Gplus-CM6
num_runs=50,
graph_loc=Gplus_loc
type_graph='SN',
num_red=1,
k=5,
dict_counter_measure = {'id': COUNTER_MEASURE_HEAR_FROM_AT_LEAST_TWO}
dict_args = {}

Lastly, to get the results in Figure 3 in the appendix, run CM4 for different values for the delay parameter:1,2,4,8