Academia.eduAcademia.edu

Figure 4 – uploaded by Andrew Crooks

See full PDFdownloadDownload figure
Fig. 4. Sample of the Resulting Synthetic Population Dataset: (A) Synthetic Individ- uals; (B) Social Network.

Figure 4 Sample of the Resulting Synthetic Population Dataset: (A) Synthetic Individ- uals; (B) Social Network.

Related Figures (13)

Table 1. Summary of the generated synthetic datasets (n differs in the size depending on the network, né[0, 14]).
Table 1. Summary of the generated synthetic datasets (n differs in the size depending on the network, né[0, 14]).
Fig. 3. Synthetic Population Results: (A) All Synthetic Individuals; (B) Household, Workplace and Educational locations within one Census Tract; (C) Agents social net- works within one Census Tract.
Fig. 3. Synthetic Population Results: (A) All Synthetic Individuals; (B) Household, Workplace and Educational locations within one Census Tract; (C) Agents social net- works within one Census Tract.
graphic location changes can also be captured based on the simulation time step. In addition, the model directs the agents to change the interaction grour based on their geographic locations. For example, if one agent is employed, then this agent will change its location from home to workplace during the daytime and this agent will interact with the agents that are within the same work social network. Due to the changes in the geographic locations, many types of dynam- ics can be captured and analyzed by using our pipeline, such as capturing social network changes or formation, tracing location changes and conducting contact tracing of disease outbreaks.
graphic location changes can also be captured based on the simulation time step. In addition, the model directs the agents to change the interaction grour based on their geographic locations. For example, if one agent is employed, then this agent will change its location from home to workplace during the daytime and this agent will interact with the agents that are within the same work social network. Due to the changes in the geographic locations, many types of dynam- ics can be captured and analyzed by using our pipeline, such as capturing social network changes or formation, tracing location changes and conducting contact tracing of disease outbreaks.
Commuting is a feature of our daily lives, which refers to going to and from work. Through daily commuting, we can wittiness rush hour traffic and other forms of congestion. As the synthetic population has home and work locations it is relativity straightforward to use this information to initialize a commuting model. Here we just show a simple example that simulates the daily commute of residents living in one census tract from Erie County of New York State and captures their location changes through the simulation (i.e. commuting, at home, at work or at an educational site). Figure6 shows the structure of the model, the process is that the mode takes the synthetic population dataset as the input to calculate the commut« path using an A* algorithm based on home location and daytime location (e.g. work). The model moves the agents on the physical road network based on th travel speed and the commute path. One step in this model represents 5 min in which the model updates the agent location on the physical road networl and location statuses. At the starting point of the model, all agents’ locatior status are at home, which we assume the model starts at 5:55 am. When it is the turn for a specific agent to commute, this agent’s location status will chang¢ to “commuting”. To determine the time for the agent to start commuting, we randomly assign a time using a normal distribution between 6 am to 9 am fo: agents to start their morning commute. When the agents reach their daytime ocations (e.g., work), their location status will change to “at daytime location” n addition, we assume all agents spend eight hours at their daytime location thus, the afternoon commute will start eight hours after the agents reach thei daytime locations. Similarly, when agents reach home, their location status wil
Commuting is a feature of our daily lives, which refers to going to and from work. Through daily commuting, we can wittiness rush hour traffic and other forms of congestion. As the synthetic population has home and work locations it is relativity straightforward to use this information to initialize a commuting model. Here we just show a simple example that simulates the daily commute of residents living in one census tract from Erie County of New York State and captures their location changes through the simulation (i.e. commuting, at home, at work or at an educational site). Figure6 shows the structure of the model, the process is that the mode takes the synthetic population dataset as the input to calculate the commut« path using an A* algorithm based on home location and daytime location (e.g. work). The model moves the agents on the physical road network based on th travel speed and the commute path. One step in this model represents 5 min in which the model updates the agent location on the physical road networl and location statuses. At the starting point of the model, all agents’ locatior status are at home, which we assume the model starts at 5:55 am. When it is the turn for a specific agent to commute, this agent’s location status will chang¢ to “commuting”. To determine the time for the agent to start commuting, we randomly assign a time using a normal distribution between 6 am to 9 am fo: agents to start their morning commute. When the agents reach their daytime ocations (e.g., work), their location status will change to “at daytime location” n addition, we assume all agents spend eight hours at their daytime location thus, the afternoon commute will start eight hours after the agents reach thei daytime locations. Similarly, when agents reach home, their location status wil
Fig. 8. Large Scale Disease Spread Model Structure.
Fig. 8. Large Scale Disease Spread Model Structure.
Fig. 9. Disease Model with Contact Tracing via Social Networks.
Fig. 9. Disease Model with Contact Tracing via Social Networks.
Fig. 10. Disease Dynamics for Two Diseases.
Fig. 10. Disease Dynamics for Two Diseases.
Fig. 11. Vaccination Opinion Dynamic Model.
Fig. 11. Vaccination Opinion Dynamic Model.
Related topics:
Agent-based modelingGeographic Information Systems (GIS)
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved