Skip to content

Evolution simulator is a desktop java application creted for the Object-Oriented Programming course at the AGH UST in 2022/2023.

Notifications You must be signed in to change notification settings

bjam24/evolution-simulator

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

14 Commits
 
 
 
 
 
 

Repository files navigation

Evolution simulator

This project was made for the Object-Oriented Programming course at the AGH UST in 2022/2023.

Description

This project is the implementation of a simple simulator based on Darwin's Theory. It is based on Multithreading so there is possibility to run many simulations at the same time. Animals are moving on the map, eating grass, and reproducing. Each move requires some energy, so animals need to eat to survive. Also, breeding consumes part of the parent's energy. An animal dies after having run out of energy.

Each animal has its genome. A genome is a combination of genomes inherited from an animal's parents. Genome is built of numbers from 0 to 7. Each number is correlated with the animal's preference to move in a specific direction. The more repetition of a digit, the more likely it is for the animal to move in the direction represented by this digit.

Desktop.21-03-2024.16-31-52.mp4

Simulation

Each day's simulation consists of the following sequence of steps:

  • removing dead animals from the map,
  • the turning and displacement of each animal,
  • consumption of plants on which animals have entered the fields,
  • reproduction of fed animals in the same field,
  • growing new plants on selected map fields.

A given simulation is described by a number of parameters:

  • height and width of the map,
  • map variant (explained in the section below),
  • starting number of plants,
  • energy provided by eating one plant,
  • number of plants growing each day,
  • plant growth variant (explained in the section below),
  • starting number of animals,
  • pet starting energy,
  • energy necessary to consider the animal as full (and ready to breed),
  • parents' energy used to create offspring,
  • minimum and maximum number of mutations in descendants (can be equal to 0),
  • mutation variant (explained in the section below),
  • the length of the animals' genome,
  • pet behavior variant (explained in the section below),

Variants

Certain aspects of the simulation are configurable and can significantly change its course. Some are simple numerical parameters (e.g. initial population sizes). Some of them, however, significantly modify its rules. This applies in particular to: the operation of the map, the operation of plant growth, the operation of mutations, and the behavior of animals.

In the case of a map, the key is how we handle its edges. We will implement the following variants:

  • globe - the left and right edges of the map are looped (if the animal goes beyond the left edge, it will appear on the right - and if it goes beyond the right, then on the left); the upper and lower edges of the map are the poles you cannot enter them (if the animal tries to go beyond these edges of the map, it remains in the field it was on and its direction changes to the opposite);

  • hell portal - if the animal goes beyond the edge of the map, it goes to a magical portal; its energy decreases by a certain amount (the same as in the case of the offspring's generation), and then it is teleported to a new, random place on the map. When it comes to plant growth, certain fields are strongly preferred, according to the Pareto principle. There is an 80% chance that the new plant will grow in the preferred field and only a 20% chance that it will grow in the second-class field. About 20% of all places on the map are preferred, 80% of places are considered unattractive. We implement the following variants:

  • forested equators - plants prefer the horizontal strip of fields in the central part of the map (imitating the equator and the surrounding area);

  • toxic corpses - plants prefer those fields where animals die least often - they grow in those fields where the fewest animals died during the simulation. In the case of mutations, we have two simple options:

  • complete randomness - mutation changes a gene to any other gene;

  • slight correction - a mutation changes a gene by 1 up or down (e.g. gene 3 can be changed to 2 or 4, and gene 0 to 1 or 7). The behavioral variants are similarly simple:

  • full predestination - the animal always executes the genes sequentially, one after the other;

  • a bit of madness - in 80% of cases, after executing a gene, the animal activates the gene immediately following it, but in 20% of cases it jumps to another, random gene.

FAQ

  • The newly born (or generated) pet is oriented in a random direction. It also has a random gene activated (not necessarily the first one).

  • The born child appears in the same field as its parents.

  • Not all parameter values must be allowed. It is better to limit the permissible ranges (especially to those that will not immediately cause the application to crash) - and inform the user that his configuration is not correct if he goes beyond them.

  • We treat energy as an integer. However, we make sure that its only source are plants (after reproduction, the sum of the energy of organisms in a given field should be the same as before reproduction).

  • If several animals compete for a plant (or for the opportunity to breed) in one field, the conflict is resolved in the following way: organisms with the highest energy are given priority,

if this does not make it possible to decide, the oldest organisms have priority,

if this does not make it possible to decide, then the organisms with the largest number of children are given priority,

If this does not allow us to decide, we choose among the tied organisms in any way we like.

  • Plants can grow where animals stand. Eating takes place at a specific point in the "circadian cycle". Then the animal no longer disturbs the existence of the plant.

  • New plants do not appear if there is no longer room for them on the map.

About

Evolution simulator is a desktop java application creted for the Object-Oriented Programming course at the AGH UST in 2022/2023.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages