Research Ideas

Homework 6, Problem 3: The Connect Four Player class
[30 extra points; individual or pair]

Submission: Submit your hw6pr3.py file to Canvas

You should copy your hw6pr2.py file to hw6pr3.py .

The Player class -- a preview

This problem asks you to write another connect-four-related class named Player.

You should write this new class in a copy of your hw6pr2.py file -- name the new copy hw6pr3.py.

In this problem, you will need to create a class named Player that evaluates connect-four boards and decides where to move next. The basic approach is the following:

Look at each column on the board. Give each column a numeric score:

-1.0 represents a column that is full (so no move can be made there)
represents a column that, if chosen as the next move, will result in a loss for the the player.
50.0 represents a column that, if chosen as the next move, will not produce a win nor produce a loss for the player (at least not in the near-term future...)
100.0 represents a column that, if chosen as the next move, will result in a win for the the player.

After obtaining a list of scores in the above format, one score per column, the computer player will choose a move by finding the column with the maximum score and playing there. If there are ties (and there will be), one of these tie-breaking strategies is used:

'LEFT' pick the leftmost high score
'RIGHT' pick the rightmost high score
'RANDOM' pick one of the high scores randomly

The more detailed descriptions below will provide a skeleton and a couple of hints for the design of your Player class and how to test it.

The Player class

Your Player class should have at least these three data members:

A one-character string representing the checker, either 'X' or 'O', being used by the connect-four Player. Warning!: remember that 'O' is capital-o, not zero. One reasonable name for this data member might be self.ox .

A string, either 'LEFT', 'RIGHT', or 'RANDOM', representing the tiebreaking type of the player. This is the name for one of the three strategies described above. One reasonable name for this data member might be self.tbt (for tiebreaking type).

A nonnegative integer representing how many moves into the future the player will look in order to evaluate possible moves. One reasonable name for this data member might be self.ply because one turn of gameplay is sometimes called a ply.

Methods required for the Player class :

You should provide your Player class with write the following methods. Be sure to try the hints on how to test each one after writing it!

__init__

__init__(self, ox, tbt, ply):

This is a constructor for Player objects that takes three arguments. (Remember that self refers to the object being constructed and that it is not explicitly passed into the constructor.) This constructor first takes in a one-character string ox: this will be either 'X' or 'O'. Second, it takes in tbt, a string representing the tiebreaking type of the player. It will be one of 'LEFT', 'RIGHT', or 'RANDOM'. The third input ply will be a nonnegative integer representing the number of moves that the player should look into the future when evaluating where to go next.

Inside the constructor, you should set the values of the data members of the object. There's not much else to do. (See below for help on the code)

__repr__

__repr__(self):

This method returns a string representing the Player object that calls it. This should simply print the three important characteristics of the object: its checker string, its tiebreaking type, and its ply. Since we will go over these two methods in class, feel free to use that code (provided here):

class Player:

""" an AI player for Connect Four """

def __init__( self, ox, tbt, ply ):

""" the constructor """

self.ox = ox

self.tbt = tbt

self.ply = ply

def __repr__( self ):

""" creates an appropriate string """

s = "Player for " + self.ox + "\n"

s += " with tiebreak type: " + self.tbt + "\n"

s += " and ply == " + str(self.ply) + "\n\n"

return s

Testing __init__ and __repr__:

>>> p = Player('X', 'LEFT', 2)

>>> p

Player for X

with tiebreak: LEFT

and ply == 2

>>> p = Player('O', 'RANDOM', 0)

>>> p

Player for O

with tiebreak: RANDOM

and ply == 0

Admittedly, testing at this point is mostly to familiarize yourself with objects of type Player.

oppCh

oppCh(self):

This method should return the other kind of checker or playing piece, i.e., the piece being played by self's opponent. In particular, if self is playing 'X', this method returns 'O' and vice-versa. Just be sure to stick with capital-O! This method is easy to test:

>>> p = Player('X', 'LEFT', 3)

>>> p.oppCh()

'O'

>>> Player('O', 'LEFT', 0).oppCh()

'X'

scoreBoard

scoreBoard(self, b):

This method should return a single float value representing the score of the input b, which you may assume will be an object of type Board. This should return 100.0 if the board b is a win for self. It should return 50.0 if it is neither a win nor a loss for self, and it should return 0.0 if it is a loss for self (i.e., the opponent won).

Testing scoreBoard
You should test all three possible output scores -- here is an example of how to test the first case. For easy copy-and-paste, many statements are in one large line:

>>> b = Board(7,6)

>>> b.setBoard( '01020305' )

>>> b

| | | | | | | |

|X| | | | | | |

|X|O|O|O| |O| |

---------------

0 1 2 3 4 5 6

>>> p = Player( 'X', 'LEFT', 0 )

>>> p.scoreBoard(b)

100.0

>>> Player('O', 'LEFT', 0).scoreBoard(b)

0.0

>>> Player('O', 'LEFT', 0).scoreBoard( Board(7,6) )

50.0

The last two examples are using objects that have not yet been assigned variable names.

Note that the tiebreak type does not affect this method at all. You can save time by having scoreBoard use the winsFor method. Recall that winsFor is in the Board class, however.

tiebreakMove

tiebreakMove(self, scores):

This method takes in scores, which will be a nonempty list of floating-point numbers. If there is only one highest score in that scores list, this method should return its COLUMN number, not the actual score! Note that the column number is the same as the index into the list scores. If there is more than one highest score because of a tie, this method should return the COLUMN number of the highest score appropriate to the player's tiebreaking type.

Thus, if the tiebreaking type is 'LEFT', then tiebreakMove should return the column of the leftmost highest score (not the score itself). If the tiebreaking type is 'RIGHT', then tiebreakMove should return the column of the rightmost highest score (not the score itself). And if the tiebreaking type is 'RANDOM', then tiebreakMove should return the column of the a randomly-chosen highest score (yet again, not the score itself).

Testing tiebreakMove
You should test for all three tiebreaking types. Here are two tests:

>>> scores = [0, 0, 50, 0, 50, 50, 0]

>>> p = Player('X', 'LEFT', 1)

>>> p2 = Player('X', 'RIGHT', 1)

>>> p.tiebreakMove(scores)

2

>>> p2.tiebreakMove(scores)

5

A hint on tiebreakMove: it's helpful to find the max of the list first (with Python's built-in max function) and then search for the column in which the max is located -- starting from an appropriate initial position -- and the self.tbt string is what determines the appropriate initial position from which to start that search.

scoresFor

scoresFor(self, b):

This method is the heart of the Player class! Its job is to return a list of scores, with the cth score representing the "goodness" of the input board after the player moves to column c. And, "goodness" is measured by what happens in the game after self.ply moves.

Admittedly, that is a lot to handle! So, here is a breakdown:

(Below is just one way to solve the problem. You do NOT have to follow it! One thing is for sure: you should use recursion.)

First, this method creates a list of all zeros (I called it scores) with length equal to the number of columns in the board b. Remember that you can use list multiplication: [0]*b.width.

Then, this method loops over all possible columns.

Base Case If a particular column is full, it assigns a -1.0 score for that column.

Another Base Case Next, if the object's ply is 0, no move is made. What's more, the column, if not full, is evaluated for the self player. (Which method in the Player class will do this?) When self.ply is 0, this means that all of the non-full columns will have the same score. After all, this is to be expected if the player is not looking at all into the future.

And Another Base Case If the game is already over, then there's no point in making any additional moves (indeed, it's not allowed!) -- simply evaluate the board and use that score for the current column under consideration.

Recursive Case But, if the object's ply is greater than 0 and the game isn't over, the code should make a move into the column that is under consideration. This will use some of the methods in Board, such as addMove.

In this case, it first checks if making that one move -- the first one into the current column under consideration -- will win the game for self. If so, that column should score 100.0! Similarly, if it does not win, but does fill up the board, it scores 50.0.

If it's not a win or a tie, however, the code next figures out what scores an opponent would give the resulting board. This means creating an opponent (which will be of Player class!) You should assume that this to-be-constructed opponent player has the same tiebreaking-style as you do -- that is, as self does. The scores reported by the opponent are NOT the scores that you should use. Rather, you will want to compute self's evaluation of the board based on the list of opponent's scores. Then, assign your score to the value of the current column's move.

Be sure to delete the checker that had been placed throughout the evaluation of this particular column.

Once all of the possible moves have been evaluated, the scoresFor method should return the complete list of scores, one per column. Typically there will be seven numbers in the list returned.

Testing scoresFor
Here is a case that will test almost all of your scoresFor method, the commands to set up the board have been placed on one line for easy copy-and-paste into your Python window:

>>> b = Board(7,6)

>>> b.setBoard( '1211244445' )

>>> b

| | | | | | | |

| | | | |X| | |

| |O| | |O| | |

| |X|X| |X| | |

| |X|O| |O|O| |

---------------

0 1 2 3 4 5 6

# 0-ply lookahead doesn't see threats...

>>> Player('X', 'LEFT', 0).scoresFor(b)

[50.0, 50.0, 50.0, 50.0, 50.0, 50.0, 50.0]

# 1-ply lookahead sees immediate wins

# (if only it were 'O's turn!)

>>> Player('O', 'LEFT', 1).scoresFor(b)

[50.0, 50.0, 50.0, 100.0, 50.0, 50.0, 50.0]

# 2-ply lookahead sees possible losses

# ('X' better go to column 3...)

>>> Player('X', 'LEFT', 2).scoresFor(b)

[0.0, 0.0, 0.0, 50.0, 0.0, 0.0, 0.0]

# 3-ply lookahead sees set-up wins

# ('X' sees that col 3 is a win!)

>>> Player('X', 'LEFT', 3).scoresFor(b)

[0.0, 0.0, 0.0, 100.0, 0.0, 0.0, 0.0]

# At 3-ply, 'O' does not see any danger

# if it moves to columns on either side...

>>> Player('O', 'LEFT', 3).scoresFor(b)

[50.0, 50.0, 50.0, 100.0, 50.0, 50.0, 50.0]

# But at 4-ply, 'O' does see the danger!

# again, too bad it's not 'O's turn...

>>> Player('O', 'LEFT', 4).scoresFor(b)

[0.0, 0.0, 0.0, 100.0, 0.0, 0.0, 0.0]

This last test may take a few seconds, even on a fast computer... .

nextMove

nextMove(self, b):

This method takes in b, an object of type Board and returns an integer -- namely, the column number that the calling object (of class Player) chooses to move to. This is the primary interface to Player, but it is really just a "wrapper" for the heavy lifting done by the other methods, particularly scoresFor. Thus, nextMove should use scoresFor and tiebreakMove to return its move.

Testing nextMove
This is similar to the previous example; again, the <<< prompts have been removed at the top for easy copy-and-paste into your Python window:

>>> b = Board(7,6);

>>> b.setBoard( '1211244445' )

>>> b

| | | | | | | |

| | | | |X| | |

| |O| | |O| | |

| |X|X| |X| | |

| |X|O| |O|O| |

---------------

0 1 2 3 4 5 6

>>> Player('X', 'LEFT', 1).nextMove(b)

0

>>> Player('X', 'RIGHT', 1).nextMove(b)

6

>>> Player('X', 'LEFT', 2).nextMove(b)

3

# the tiebreak does not matter

# if there is only one best move...

>>> Player('X', 'RIGHT', 2).nextMove(b)

3

# again, the tiebreak does not matter

# if there is only one best move...

>>> Player('X', 'RANDOM', 2).nextMove(b)

3

Putting it all together: Board's playGame method

playGame

Add the following method to your Board class in hw6pr3.py ! (Do not change hw6pr2.py.)

playGame(self, px, po):
Copy and change your hostGame method from hw6pr2 in order to create this playGame method. playGame does just that -- it calls on the nextMove method in px and po, which will be objects of type Player in order to play a game.

One additional twist: you should handle the case in which either px or po is the string 'human' instead of an object of type Player. In this 'human' case, the playGame method should simply puase and ask the user to input the next column to move for that player, with error-checking just as in hostGame.

See below for some determininstic -- that is, non-random -- examples that you can use to test your playGame method:

Testing playGame
Some deterministic examples:

>>> px = Player('X', 'LEFT', 0)

>>> po = Player('O', 'LEFT', 0)

>>> b = Board(7,6)

>>> b.playGame(px, po)

# Lots of boards omitted...

|O|O|O| | | | |

|X|X|X| | | | |

|O|O|O| | | | |

|X|X|X| | | | |

|O|O|O| | | | |

|X|X|X|X| | | |

---------------

0 1 2 3 4 5 6

X wins!

Example #2 (notice the game ends faster!):

>>> px = Player('X', 'LEFT', 1)

>>> po = Player('O', 'LEFT', 1)

>>> b = Board(7,6)

>>> b.playGame(px, po)

# Lots of boards omitted...

|O|O| | | | | |

|X|X| | | | | |

|O|O| | | | | |

|X|X| | | | | |

|O|O|O| | | | |

|X|X|X|X| | | |

---------------

0 1 2 3 4 5 6

X wins!

If you have gotten to this point, you have completed problem 3! You should submit your hw6pr3.py file at Canvas .

Links

Lab 6

Homework 6