Ram Prasad

VU22EECE0100102

Wumpus World

Aim: To Implement WUMPUS World using Python.

Apparatus: pc, Google COLAB
Introduction:
This document records an experiment where an agent navigates through a virtual world
represented by a 4x4 grid. The agent's goal is to explore the environment, avoid dangers, find
gold, and return to the starting position.

Python Code :

def learnagent(world,i,j):
'''Function for an agent to know what position contains which
environment objects'''
if (world[i][j]==9):
agi,agj=i,j
print("\nNow the agent is at "+str(agi)+","+str(agj))
print("You came across a stench")
return agi,agj

elif (world[i][j]==7):
agi,agj=i,j
print("\nNow the agent is at "+str(agi)+","+str(agj))
print("You came across a pit")
return -5,-5

elif (world[i][j]==6):
agi,agj=i,j
print("\nNow the agent is at "+str(agi)+","+str(agj))
print("You found gold")
return -4,-4

elif (world[i][j]==5):
agi,agj=i,j
print("\nNow the agent is at "+str(agi)+","+str(agj))
print("You feel breeze")
return agi,agj

elif (world[i][j]==-1):
agi,agj=i,j
 print("\nNow the agent is at "+str(agi)+","+str(agj))
print("You met wumpus")
return -5,-5
else: #if world environment was empty
agi,agj=i,j
print("\nNow the agent is at "+str(agi)+","+str(agj))
return agi,agj

def checkinp(agi,agj):
'''Function for checking input going in forward direction to get gold'''
if(agi==0 and agj==0):
print("\nyou can go at "+str(agi+1)+" "+str(agj)) #can move
upward
print("you can go at "+str(agi)+" "+str(agj+1)) #can move right
agvi=int(input("\nEnter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi+1 and agvj==agj or agvi==agi and agvj==agj+1):
return agvi,agvj
else:
return -5
elif(agi==3 and agj==0):
print("\nyou can go at "+str(agi-1)+" "+str(agj)) #can go left
print("you can go at "+str(agi)+" "+str(agj+1)) #can go right
agvi=int(input("\nEnter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi-1 and agvj==agj or agvi==agi and agvj==agj+1):
return agvi,agvj
else:
return -5
elif(agi==3 and agj==3):
print("\nyou can go at "+str(agi-1)+" "+str(agj)) #can go down
print("you can go at "+str(agi)+" "+str(agj-1)) #can go left
agvi=int(input("\nEnter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi-1 and agvj==agj or agvi==agi and agvj==agj-1):
return agvi,agvj
else:
return -5
elif(agi==0 and agj==3):
print("\nyou can go at "+str(agi+1)+" "+str(agj)) #can go upward
 print("you can go at "+str(agi)+" "+str(agj-1)) #can go left
agvi=int(input("\nEnter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi+1 and agvj==agj or agvi==agi and agvj==agj-1):
return agvi,agvj
else:
return -5,-5
elif(agi==1 and agj==0 or agi==2 and agj==0 or agi==3 and agj==0):
print("\nyou can go at "+str(agi+1)+" "+str(agj)) #can go upward
print("you can go at "+str(agi)+" "+str(agj+1)) #can move right
agvi=int(input("\nEnter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi+1 and agvj==agj or agvi==agi and agvj==agj+1):
return agvi,agvj
else:
return -5,-5
elif(agi==0 and agj==3 or agi==1 and agj==3 or agi==2 and agj==3 or
agi==3 and agj==3):
print("you can go at "+str(agi+1)+" "+str(agj)) #can go upward
print("you can go at "+str(agi)+" "+str(agj-1)) #can go left
agvi=int(input("Enter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi+1 and agvj==agj or agvi==agi and agvj==agj-1):
return agvi,agvj
else:
return -5,-5
elif(agi==3 and agj==1 or agi==3 and agj==2 or agi==3 and agj==3):
print("\nyou can go at "+str(agi)+" "+str(agj+1)) #can go right
print("you can go at "+str(agi)+" "+str(agj-1)) #can go left
print("you can go at "+str(agi-1)+" "+str(agj)) #can move downward
agvi=int(input("\nEnter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi and agvj==agj+1 or agvi==agi and agvj==agj-1 or
agvi==agi-1 and agvj==agj):
return agvi,agvj
else:
return -5,-5
else:
print("\nyou can go at "+str(agi)+" "+str(agj+1)) #can go right
print("you can go at "+str(agi)+" "+str(agj-1)) #can go left
 print("you can go at "+str(agi+1)+" "+str(agj)) #can move upward
agvi=int(input("\nEnter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi and agvj==agj+1 or agvi==agi and agvj==agj-1 or
agvi==agi+1 and agvj==agj):
return agvi,agvj
else:
return -5,-5

def checkinpreverse(agi,agj):
'''Function for checking input going in reverse direction to get back to
original position'''
if(agi==0 and agj==3):
print("you can go at "+str(agi)+" "+str(agj-1)) #can go left
agvi=int(input("\nEnter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi and agvj==agj-1):
return agvi,agvj
else:
return -5,-5
elif(agi==0 and agj==2 or agi==0 and agj==1):
print("you can go at "+str(agi)+" "+str(agj+1)) #can go right
print("you can go at "+str(agi)+" "+str(agj-1)) #can go left
agvi=int(input("\nEnter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi and agvj==agj-1 or agvi==agi and agvj==agj+1 ):
return agvi,agvj
else:
return -5,-5
elif(agi==1 and agj==0 or agi==2 and agj==0):
print("\nyou can go at "+str(agi-1)+" "+str(agj)) #can go downward
print("you can go at "+str(agi)+" "+str(agj+1)) #can move right
agvi=int(input("\nEnter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi-1 and agvj==agj or agvi==agi and agvj==agj+1):
return agvi,agvj
else:
return -5,-5
elif(agi==1 and agj==3 or agi==2 and agj==3):
print("you can go at "+str(agi-1)+" "+str(agj)) #can go downward
 print("you can go at "+str(agi)+" "+str(agj-1)) #can go left
agvi=int(input("Enter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi-1 and agvj==agj or agvi==agi and agvj==agj-1):
return agvi,agvj
else:
return -5,-5
else:
print("\nyou can go at "+str(agi-1)+" "+str(agj)) #can go downward
print("you can go at "+str(agi)+" "+str(agj-1)) #can go left
print("you can go at "+str(agi)+" "+str(agj+1)) #can go right
agvi=int(input("\nEnter input for row => "))
agvj=int(input("Enter input for column => "))
if(agvi==agi-1 and agvj==agj or agvi==agi and agvj==agj-1 or agvi==agi
and agvj==agj+1):
return agvi,agvj
else:
return -5,-5

world=[ [0,5,7,5],
[9,0,5,0],
[-1,6,7,5],
[9,0,5,7] ] #declaration of a world

agi,agj=0,0 #initial agent position

print("\n\n\ninitially agent is at "+str(agi)+","+str(agj))
print("\nyou can go at "+str(agi+1)+" "+str(agj))
print("you can go at "+str(agi)+" "+str(agj+1))

agvi=int(input("Enter input for row => "))

agvj=int(input("Enter input for column => ")) #taking row and column
values
if(agvi==1 and agvj==0 or agvi==0 and agvj==1):
agi,agj=learnagent(world,agvi,agvj) #if input valid calling learn
agent function
else:
print("Not valid")

while(agi>=0):
agvi,agvj=checkinp(agi,agj)
 if(agvi!=-5 and agvj!=-5):
agi,agj=learnagent(world,agvi,agvj)
else:
print("\nNot valid")

if(agi==-5):
print("\nGame over Sorry try next time!!!")
else:
print("\nYou have unlocked next level move back to your initial
position") #acquired gold

agi,agj=2,1 #implementation of reverse logic

while(agi>=0):
agvi,agvj=checkinpreverse(agi,agj)
if(agvi==0 and agvj==0):
agi,agj=-4,-4
elif(agvi!=-5 and agvj!=-5):
agi,agj=learnagent(world,agvi,agvj)
else:
print("\nNot valid")

if(agi==-5):
print("\nYou were really close but unfortunately you failed!!! Try
next time")
else:
print("\nHurray You won!!!!! Three cheers.")
Environment:
● A 4*4 grid of rooms.
● The agent initially in room square [0, 0], facing toward the right.
● Location of Wumpus and gold are chosen randomly except the first square [0,0].
● Each square of the cave can be a pit with probability 0.2 except the first square.
● Agent can Move:
1. Left turn,
2. Right turn
3. Move forward
4. Grab

Note: Diagonal movement is not allowed.

Observations:

initially agent is at 0,0

you can go at 1 0
you can go at 0 1
Enter input for row => 1
Enter input for column => 0

Now the agent is at 1,0

You came across a stench

you can go at 2 0
you can go at 1 1

Enter input for row => 1

Enter input for column => 1

Now the agent is at 1,1

you can go at 1 2
you can go at 1 0
you can go at 2 1

Enter input for row => 2

Enter input for column => 1

Now the agent is at 2,1

You found gold

You have unlocked next level move back to your initial position

you can go at 1 1
you can go at 2 0
you can go at 2 2

Enter input for row => 1

Enter input for column => 1

Now the agent is at 1,1

you can go at 0 1
you can go at 1 0
you can go at 1 2

Enter input for row => 1

Enter input for column => 0

Now the agent is at 1,0

You came across a stench

you can go at 0 0
you can go at 1 1

Enter input for row => 1

Enter input for column => 1

Now the agent is at 1,1

you can go at 0 1
you can go at 1 0
you can go at 1 2

Enter input for row => 1

Enter input for column => 0

Now the agent is at 1,0

You came across a stench

you can go at 0 0
you can go at 1 1

Enter input for row => 0

Enter input for column => 0

Hurray You won!!!!! Three cheers.

Explanation:

Phase 1: Moving Forward

● Initial Movement Options:

○ The agent can move right or down from (0, 0).
● Agent's Movement:
○ User inputs to move to (1, 0).
● Encounter:
○ The agent encounters a stench at (1, 0).
● Learning and Decision:
○ Agent's new position is recorded, and it identifies the stench.
● Continued Exploration:
○ Agent continues exploring based on available movements and encounters.

Phase 2: Collecting Gold and Reversing

1. Gold Collection:
○ Agent collects gold at (1, 2).
2. Return Journey:
○ Agent needs to return to (0, 0).
3. Completion:
○ Agent successfully returns to (0, 0).

Result: The experiment demonstrated the agent's ability to navigate a complex

environment, avoid dangers, achieve goals, and return to the starting position. This exercise
highlighted the effectiveness of the agent's decision-making process and pathfinding
capabilities.