Opinion dynamics simulation from real world data 1.0.0
This is an opinion dynamics simulation using Jupiter notebook with python. In the code agents interact in random pairs (as in a fully connected graph) and agent 1 changes her opinion depending on the opinion of agent 2.
The code uses three main effects:
a) Social influence (i.e. agent 1 moves in the direction of agent 2)
b) Flipping (i.e. agent 1 has a certain probability of changing her agreement level, while preserving her certainty)
c) Certainty-induced noise (i.e. the opinion of the agent has some random fluctuations whose amplitude depends on the initial
The values of these rules are determined from experiments with participants.