D. Events

Single Events

The Event Panels control the events that can be executed inside the simulation. Some events are applied to agents, and others to institutions. Events, when initiated, are added to an internal queue and will be executed during the next Checkpoint.

Any event can be configured, i.e. various parameters can be set for it in order to affect only certain parts of the world or only certain agents. This is done with the Configure Button of each individual Event panel and will be covered in-depth in the Configure Events subsection of this chapter. For now, we will proceed working with a uniform distribution of the event's effects across all parameters.

There are six event panels in total. Of these six, the first four contain two similar types of events. We will now cover all six in-depth:

• Institutional structure removal affects the associations between agents and institutions. There are two types:

• Apostasy: A number of agents abandon their institutions. Internally, these agents will now be assigned institutions with empty traits. A change in the distribution parameter of the event affects the agents.

• Destruction: A number of institutions are destroyed. The agents that belonged to them are each assigned a new institution with empty traits. The change in the distribution parameter of the event affects the institutions.

• Institutional content removal removes traits inside the institutions. Again, there are two types:

• Partial: Some traits are removed from a number of institutions. A change in the distribution parameter of this event affects the institution's traits. When a Non Probablistic Distribution (see Configure Events Section) is used for this event, then there is no difference between this event (Partial) and the next (Full).

• Full: All traits are removed from a number of institutions. A change in the distribution parameter of this event affects all the traits of the institutions.

• Institutional conversion instroduces foreigner (invader) traits into institutions. This invader trait is new to the population and different from any of the previously existent traits. This event groups two types:

• Partial: Some traits from some institutions are converted into foreigner (invader) traits. A change in the distribution parameter of the event affects some traits of the institutions. When a Non Probablistic Distribution (see Configure Events Section) is used for this event, then there is no difference between this event (Partial) and the next (Full).

• Full: All traits from some institutions are converted into foreigner (invader) traits. A change in the distribution parameter of the event affects all the traits of the institutions.

• Settlement (called invasion in previous versions): A number of agents such as settlers (foreigners with their own institutions) are introduced into the simulation. They enter into positions that were occupied by other agents. Settlers have only foreigner traits in their cultural vectors, and all of them belong to the same settler institution. A change in the distribution parameter of this event affects the distribution of agents that will be replaced by settlers.

• Immigration: A number of agents such as immigrants (foreigners without their own institutions) are introduced into the simulation. They enter into positions that were occupied by other agents. Immigrants have only foreigner traits in their cultural vectors and, in principle, don't belong to any institution until other agents persuade them to do so. A change in the distribution parameter of this event affects the distribution of agents that will be replaced by settlers.

• Decimation (called genocide in previous versions): A number of agents are killed within the current population. Internally, all traits of the simulation are replaced by a dead trait. A change in the distribution parameter of this event affects the distribution of agents to be killed.

• Parameter Change Event: This event doesn't affect agents or institutions directly. Instead, with it, it is possible to change many of the parameters of the simulation that were initially set in Simulation -> Parameters. (see Initial Parameters).

Configure Events

The configuration of each event appears in blue font in each Event Panel.

Clicking the Configure Event Button opens the Configuration Distribution Panel, which enables you to change the parameters.

Changing the configuration for an event consists (except for Parameter Change Event) of adjusting the distribution allocated on the grid. In general, the distribution will indicate which agents or institutions are affected by the event. Four types of distributions are implemented. The first two are probabilistic and the latter two are deterministic.

• Probabilistic Distributions: These distributions use a probabilistic function to assign a probability to each agent or institution. This probability is used to decide whether the event affects that particular agent or institution.

• Uniform Distributions: They assign the same probability to all agents or institutions. The only parameter in this distribution is the probability of an event occurring to an agent or institution.

• Aprox. Normal Distributions: They assign probabilities to each agent or institution according to the normal distribution. The agent in the center of the distribution receives a probability of a maximum value, and the other agents receive a probability depending on the distance from this center. The first two parameters indicate in which row and column the distribution will be centered. You can use -1 to select the row and column randomly. The third parameter specifies the maximum value. The fourth parameter is the standard deviation, to indicated how much the event spreads from its center.

• Est. Normal Distributions: Equivalent to Aprox. Normal Distributions except that instead of the standard deviation (as fourth parameter), it receives a proportion of cells that will be affected. Internally, CulSim uses this proportion to estimate a corresponding standard deviation. The rest remains the same.

• Non Probabilistic Distributions: These distributions select the specific agents or institutions that will be affected by the event. The event will occur with a probability of 1.0 to the agents or institutions selected. When a Non Probabilistic Distribution is used for the Institution Content Remove and Institution Conversion there is then no difference between the Partial and Full version of their events.

• Neumann: This distribution uses Von Neuman neighborhoods with a distance r to distribute the events, to select the agents or institutions that will be affected by the event. The first two parameters indicate in which row and column the neighborhood will be centered. Use -1 to select the row and column randomly. The third parameter indicates the radius of the neighborhood, i.e. how far the event spreads from its center.

• Rectangular: These distributions use two coordinates (by providing rows and colums) to define a rectangle on the grid. The institutions or agents that fall into this rectangle are the ones that will be affected by the event. The first two parameters indicate the first coordinate of the rectangle, and the last two parameters the second coordinate of the rectangle.

Parameter Change Event

A special case for the configuration of events is the Parameter Change Event. Select new parameters that you want to apply to the simulation here. When the Configure Event button for this event is pressed, the following dialog box is shown:

It is possible to change most of the parameters that were set in the Initial Parameter Setup, except those that involve changes in static data structures (arrays) such as the size of the grid, the cultural vector, or the neighbors.

Composed Events

You can create combinations of events with the Add Event Button . The events are added to the Event Set:

The order in which the events are added matters. For example, a decimation after a settlement will kill some of the settlers, but this would not happen if the settlement came after the decimation. It is also possible to add two events of the same type. This is useful, for example to simulate a settlement occurring at two different locations of the grid. You can start a new event set by cleaning the list with the Clean Event Set Button .

The composed events can be saved or loaded to files with the Save Event Button . You can apply the same event to different simulation states or configurations. Moreover, saved events are essential when using Batch Mode.