Agentbased computational economics

1
Agent-based computational economics

• Michael Neugart
• Session 4

2
Agenda

• Managing lots of agents using ArrayList
• Inheritance
• Useful things
• Sorting
• Generating random numbers

3
Lists

• Java features Collections which are data
  structures holding objects
• A type of Collections is ArrayList
• We will use ArrayList to give our agents a home
• You can create an ArrayList by typing
• ArrayList list new ArrayList()

4
Lists

• public class Worker
• //what follows are the instance variables
• int id
• double humanCapital
• double wage
• double random
• //what follows is the constructor
• public Worker (int i, double h, double w)
• id 0
• humanCapital h
• wage w
• //what follows are the methods
• public void luck()
• random getNextDoubleFromTo(1,2)
• wage random humanCapital

5
Lists

• public class exampleLists
• //what follows calls the main method
• public static void main(String args)
• int n 5
• double random
• //creates the list
• ArrayList workerList new ArrayList()
• //fills the list with workers
• for(int i 0 i lt n i)
• random getNextDoubleFromTo(1000,2000)
• Worker aWorker new Worker(i,random,1500)
• workerList.add(aWorker)
• //continues on next slide

6
Lists

• //this prints the worker ids
• for(int i0 i lt workerList.size() i)
• Worker aWorker (Worker) workerList.get(i)
• System.out.println(aWorker.id)
• //looks if list is empty
• if(workerList.isEmpty())
• System.out.println(workerList is empty
• else
• System.out.println(workerList is not
  empty
• //removes the third object from the workerList
• Worker aWorker (Worker) workerList.get(2)
• workerList.remove(aWorker)
• //checks if a workerList contains a certain
  object
• if(!workerList.contains(aWorker))
• System.out.println(does not contain
  aWorker)

7
Lists

• Exercises for you
• Calculate the maximum human capital among the
  workers
• Calculate average human capital among the workers
• Sort the list of workers in ascending order of
  workers human capital. (Hint Create a temporary
  additional list. You can use list2 list1 to
  write the contents of list1 into list2.)

8
Lists

• There are several predefined routines you may use
  when working with lists. Quite helpful is
• Collections.shuffle() which randomizes the
  order of objects in a list.
• There is an alternative for going through lists.
  You may define an iterator
• it workerList.iterator()
• while(it.hasNext())
• Worker aWorker (Worker)it.next()
• System.out.println(aWorker.id)

9
Lists

• You can also concatenate two lists. Say
  workerList is one of these lists and otherList
  shall be merged with it. Then you may do the
  following
• ArrayList joinedList new ArrayList(workerList)
• joinedList.addAll(otherList)

10
Inheritance

• One of the three principles of Java programming
• You can create a general class (superclass). This
  class can then be inherited by more specific
  classes (subclass).
• In the specific class you add things that are
  unique to it
• The general declaration is
• Class subclass-name extends superclass-name
• //body of class

11
Inheritance

• Here is an example taken from Schildts book, p.
  252
• First a superclass is defined for two-dimensional
  objects.
• Then a subclass is added defining triangles.
• Finally a little program determines the
  properties of some examples of triangles.

12
Inheritance

• // A simple class hierarchy.
• // A class for two-dimensional objects.
• class TwoDShape
• double width
• double height
• void showDim()
• System.out.println("Width and height are "
• width " and " height)
  

13
Inheritance

• // A subclass of TwoDShape for triangles.
• class Triangle extends TwoDShape
• String style
• double area()
• return width height / 2
• void showStyle()
• System.out.println("Triangle is " style)

14
Inheritance

• class Shapes
• public static void main(String args)
• Triangle t1 new Triangle()
• Triangle t2 new Triangle()
• t1.width 4.0
• t1.height 4.0
• t1.style "isosceles"
• t2.width 8.0
• t2.height 12.0
• t2.style "right"
• System.out.println("Info for t1 ")
• t1.showStyle()
• t1.showDim()
• System.out.println("Area is " t1.area())
• System.out.println()

15
Inheritance

• Note, that the Triangle subclass adds style,
  area() and showStyle, and in addition it includes
  all the features of the TwoDShape class.

16
A solution to sorting

• // Sorting workerList
• ArrayList temp new ArrayList()
• while (!workerList.isEmpty())
• Worker poorest (Worker)workerList.get(0)
• for (int i0 iltworkerList.size() i)
• Worker aWorker (Worker)workerList.get(i)
  
• if (aWorker.getWage() lt
  poorest.getWage())
• poorest aWorker
• workerList.remove(poorest)
• temp.add(poorest)
• workerList temp
• System.out.println("Sorted list
  workerList)

17
Sorting objects in ArrayList using
Collections.sort()

• Collections.sort(list, new Comparator()  
• public int compare(Object o1, Object o2)
• Person p1 (Person) o1
• Person p2 (Person) o2
• return p1.getFirstName().compareToIgnoreCase(
  p2.getFirstName())  
• )  
• System.out.println(list)

Source http//www.albeesonline.com/blog/2008/10/1
6/sorting-an-arraylist-of-objects/
18
Generating random numbers

• We already saw in a previous example that we can
  draw a random number from a uniform distribution
• int random getNextIntFromTo(0,2)
• double random getNextDoubleFromTo(0,2)
• double random Math.random()
• Repast allows to draw random numbers from all
  kinds of distributions

19
Generating random numbers

• This is what you have to do
• Import package
• import uchicago.src.sim.util.Random
• at the beginning of your class file
• Make sure that your Eclipse project has included
  the colt.jar. Check by right-clicking on the
  project folder, go to Properties and then to Java
  Build Path
• You can draw a number from a normal distribution
  by

20
Generating random numbers

• Random.createNormal()
• double var Random.normal.nextDouble(0.5,
  0.3)
• Or you do
• Random.createNormal(0.5,0.3)
• double var Random.normal.nextDouble()
• Note that first entry denotes the mean and the
  second entry the standard deviation.

21
Generating random numbers

• Check the uchicago.src.sim.util.Random class to
  find out about the other distributions which you
  may use with Repast