Top-Down Design

Don Miner & Sue Evans

Adapted from the CS1 Course at Swarthmore College by Lisa Meeden

Press space bar for next slide

Learning Objectives

Ability to break a large problem down using top-down design
Become familiar with writing code incrementally
Realize that independent testing of functions is possible and should be done as they are written
Understand that further testing is necessary when all parts are complete and integrated.
Realize that even though you may have had working code at one time, if you've made any changes to the code, including just adding comments, it needs to be retested. This is known as regression testing.
Become familiar with python's random module

Top-Down Design

Large projects can be unmanageable to code.
Where do you start?
How do you plan out a problem?
Top-Down Design is a proven technique for tackling complex problems.

Top-Down Design

In top-down design, you break down a problem into subproblems with functions.
Then, you break these subproblems into subsubproblems, and so forth.
Eventually, you get to a place where you can start actual implementation.
This design process helps you understand the sequence of events.
It helps to segment the problem into understandable pieces.
This helps to organize your code and makes it easier to read.

Example problem: Box plot

What is a box plot ?
A box plot or box and whisker plot is a way of graphically representing statistics of numerical data by using 5-number summaries of the sample.
The 5-numbers are :
- the minimum value
- the lower quartile
- the median
- the upper quartile
- the maximum value

Example: Box plot (specification)

Lets write a program that takes a list of numbers, then prints out a box plot.

Print out a greeting to the user explaining the program.
Take a list of integers as input from either a file or the terminal.
We want to print out the box plot as well as the following statistics:
- the minimum
- the lower quartile
- the median
- the upper quartile
- the maximum
The output should look like:
```
|----[    |      ]-----|

min: 0
Q1: 4
median: 8
Q3:14
max: 19
```

Example: Box plot (first level)

The problem has 4 major parts: the greeting, getting input, calculating the statistics and printing the output.
For each of these parts, let's define a function.
For each function, think about what it should return (output) and what it should take in as parameters (input).
The only part we are ready to implement after breaking the problem up is the greeting. For the others, we will use a print statement to hold the place of our future code. These functions that aren't true implementations yet, and just contain a print statement, are called stubs.

Our code so far:

# Filename: boxplot.py
# Author:   Don Miner & Sue Evans
# Date:     8/22/09
# Section:  All
# Email:    bogar@cs.umbc.edu
#
# This program will produce a box plot for the 
# numerical data the user provides either from 
# a file or from typing at the terminal.


# printGreeting() prints an explanation of the program
# Inputs: none
# Output: none
def printGreeting():
    print "\nThis program will produce a box plot"
    print "for the numerical data you enter."
    print "Welcome to my box plot program!\n"


# getInputFromUser() prompts the user for how he will 
# give input, either from a file or from the terminal
# gets the input and returns it.
# Inputs: none
# Output: a list of numbers
def getInputFromUser():
    print "In getInputFromUser()"


# calculateStatistics() takes a list of numbers and
# calculates the minimum, the first quartile, the
# median, the third quartile and the maximum
# Inputs: a list of numbers called numbers
# Outputs: minimum, q1, median, q2, maximum
def calculateStatistics(numbers):
    print "In calculateStatistics(numbers)"
    return 1, 2, 3, 4, 5

# printBoxPlot() takes the 5 statistical values, 
# prints them and the box plot for them.
# Inputs: minimum, q1, median, q2, maximum
# Output: none
def printBoxPlot(minimum, q1, median, q3, maximum):
    print "In printBoxPlot(minimum, q1, median, q3, maximum)"


def main():

    printGreeting()
	
    numberList = getInputFromUser()

    minimum, q1, median, q3, maximum = calculateStatistics(numberList)

    printBoxPlot(minimum, q1, median, q3, maximum)

   
main()

getInputFromUser() will return a list of integers, because we have to give calculateStatistics(numberList) the data to crunch the numbers.

At this point, the progression of our problem makes sense and the pieces are clear. However, it doesn't really do anything yet. The program actually runs at this point and we'll keep it running throughout this process. main() and printGreeting() are finished.

linuxserver1.cs.umbc.edu[108] python box.py

This program will produce a box plot
for the numerical data you enter.
Welcome to my box plot program!

In getInputFromUser()
In calculateStatistics(numbers)
In printBoxPlot(minimum, q1, median, q3, maximum)
linuxserver1.cs.umbc.edu[109]

Example: Box plot (`printBoxPlot()`)

Lets expand printBoxPlot().
We don't have to do things in order!
Remember, printBoxPlot() takes 5 integer parameters, and prints output that looks like:
```
|----[    | ]--|

min: 0
Q1: 4
median: 8
Q3:9
max: 11
```
Since we are using top-down design, we can break this subproblem up, just like we did main().
First of all, we need to print the box plot, which is a task in itself.
Then, we should print out the data values. This is straightforward enough that we don't need a subproblem to do this.

We should replace the print statement in the stub of printBoxPlot() with the following code:

# printBoxPlot() takes the 5 statistical values, 
# prints them and the box plot for them.
# Inputs: minimum, q1, median, q3, maximum
# Output: none
def printBoxPlot(minimum, q1, median, q3, maximum):

   printBoxPlotGraph(minimum, q1, median, q3, maximum)

   print "min:", minimum
   print "Q1:", q1
   print "median:", median
   print "Q3:", q3
   print "max:", maximum


# printBoxPlotGraph() takes the 5 statistical values
# and prints just the box plot graph for them.
# Inputs: minimum, q1, median, q3, maximum
# Output: none
def printBoxPlotGraph(minimum, q1, median, q3, maximum):
    print " In printBoxPlotGraph(minimum, q1, median, q3, maximum)"

Example: Box plot (`printBoxPlotGraph()`)

Let's print the box plot graph now!
The length of each segment needs to be calculated to know the number of characters to use.
We use some tricky multiplication syntax with strings that allows us to repeat characters. For example:
```
>>> "hello"*3
'hellohellohello'
>>> "!"*5
'!!!!!'
```

Lets start writing code. Our subproblem will find the sizes of the pieces of the box plot, then construct each piece, then concatenate them together.

# printBoxPlotGraph() takes the 5 statistical values
# and prints just the box plot graph for them.
# Inputs: minimum, q1, median, q3, maximum
# Output: none
def printBoxPlotGraph(minimum, q1, median, q3, maximum): 

    # find the sizes of the pieces
    whiskerLeftSize = q1 - minimum 
    boxLeftSize = median - q1 
    boxRightSize = q3 - median 
    whiskerRightSize = maximum - q3 
   
    # construct the pieces
    # the space before the minimum
    blankLeft = ' ' * minimum
    whiskerLeft = '-' * whiskerLeftSize 
    boxLeft = ' ' * boxLeftSize 
    boxRight = ' ' * boxRightSize 
    whiskerRight = '-' * whiskerRightSize 
   
    # concatenate the pieces
    # break this into left & right so lines don't wrap
    left = blankLeft + '|' + whiskerLeft + '[' + boxLeft + '|' 
    right = boxRight + ']' + whiskerRight + '|' 
    boxplot = left + right 
   
    print boxplot

This almost seems like a mini problem itself! (that's because it is)
We can test this function to see if it works, without needing the rest of the problem done. Just add the following code to the end of the program instead of calling main():
```
printBoxPlotGraph(1, 3, 5, 8, 11)
```
When we run the program, we get this output, which is what we expected.
```
 |--[  |   ]---|
```
We could have broken this part down even further...
If you notice a subproblem that would make a good module, then you should definitely break the problem down further.
In this case, there is no modular subproblem, so it's up to you when to stop.

Example: Box plot

Lets take a step back and see what we have so far.

Notice that we haven't changed any of main() or the other functions.

# printGreeting() prints an explanation of the program
# Inputs: none
# Output: none
def printGreeting():
    print "\nThis program will produce a box plot"
    print "for the numerical data you enter."
    print "Welcome to my box plot program!\n"


# getInputFromUser() prompts the user for how he will 
# give input, either from a file or from the terminal
# gets the input and returns it.
# Inputs: none
# Output: a list of numbers
def getInputFromUser():
    print "In getInputFromUser()"


# calculateStatistics() takes a list of numbers and
# calculates the minimum, the first quartile, the
# median, the third quartile and the maximum
# Inputs: a list of numbers called numbers
# Outputs: minimum, q1, median, q3, maximum
def calculateStatistics(numbers):
    print "In calculateStatistics(numbers)"
    return 1, 2, 3, 4, 5


# printBoxPlot() takes the 5 statistical values, 
# prints them and the box plot for them.
# Inputs: minimum, q1, median, q3, maximum
# Output: none
def printBoxPlot(minimum, q1, median, q3, maximum):

    printBoxPlotGraph(minimum, q1, median, q3, maximum)

    print "min:", minimum
    print "Q1:", q1
    print "median:", median
    print "Q3:", q3
    print "max:", maximum


# printBoxPlotGraph() takes the 5 statistical values
# and prints just the box plot graph for them.
# Inputs: minimum, q1, median, q3, maximum
# Output: none
def printBoxPlotGraph(minimum, q1, median, q3, maximum):

    # find the sizes of the pieces
    whiskerLeftSize = q1 - minimum
    boxLeftSize = median - q1
    boxRightSize = q3 - median
    whiskerRightSize = maximum - q3
   
    # construct the pieces
    # the space before the minimum
    blankLeft = ' ' * minimum 
    whiskerLeft = '-' * whiskerLeftSize
    boxLeft = ' ' * boxLeftSize
    boxRight = ' ' * boxRightSize
    whiskerRight = '-' * whiskerRightSize
   
    # concatenate the pieces
    # break this into left & right so lines don't wrap
    left = blankLeft + '|' + whiskerLeft + '[' + boxLeft + '|'
    right = boxRight + ']' + whiskerRight + '|'
    boxplot = left + right
   
    print boxplot


def main():

   printGreeting()
	
   numberList = getInputFromUser()

   minimum, q1, median, q3, maximum = calculateStatistics(numberList)

   printBoxPlot(minimum, q1, median, q3, maximum)
   

printBoxPlotGraph(1, 3, 5, 8, 11)

Example: Box plot (`getInputFromUser()`)

Next, we will work on getInputFromUser().
This function should prompt the user for how he will give input, then get input in that way.

Then, we should return the list with the integers in it.

# getInputFromUser() prompts the user for how he will 
# give input, either from a file or from the terminal
# gets the input and returns it.
# Inputs: none
# Output: a list of numbers
def getInputFromUser():

    choice = 'invalid'
    while choice != 'file' and choice != 'type':
       choice = raw_input("Where is the input? (type 'file' or 'type') ")
   
    if choice == 'file':
        return getInputFromFile()
    elif choice == 'type':
        return getInputFromTerminal()

 
# getInputFromFile() opens the file indicated by the
# user, reads the numerical data from it, closes the
# file and returns the list of numbers.
# Inputs: none
# Output: a list of numbers
def getInputFromFile():
    print "In getInputFromFile()"


# getInputFromTerminal() gets the numerical 
# data from the keyboard as the user enters
# it and returns the list of numbers.
# Inputs: none
# Output: a list of numbers
def getInputFromTerminal():
    print "In getInputFromTerminal()"

Example: Box plot (`getInputFromTerminal()`)

Next, we'll work on getInputFromTerminal().
This function will ask how many numbers there are, then take the numbers in.

Then, we should return the list with the integers in it.

# getInputFromTerminal() gets the numerical 
# data from the keyboard as the user enters
# it and returns the list of numbers.
# Inputs: none
# Output: a list of numbers
def getInputFromTerminal():

   n = int(input("How many numbers are there? "))
   
   # initialize an empty list
   numberList = []
   
   for i in range(n):
      newNumber = int(input('> '))
      
      # add this new number to the end of the number list
      numberList.append(newNumber)

   # return the number list
   return numberList

We can test our functions by running:
print 'test:',
getInputFromUser()
instead of main():

Where is the input? (type 'file' or 'type') type
How many numbers are there? 4
> 9
> 3
> 5
> 1
test: [9, 3, 5, 1]

Success!

Lets forget about getInputFromFile() for now...

Example: Box plot (`calculateStatistics()`)

Next, we will work on calculateStatistics.
This function will take a list of integers.
We will find the statistics by sorting the list, then counting.

Then, the function returns the values needed by the next step in main.

# calculateStatistics() takes a list of numbers and
# calculates the minimum, the first quartile, the
# median, the third quartile and the maximum
# Inputs: a list of numbers called numbers
# Outputs: minimum, q1, median, q3, maximum
def calculateStatistics(numbers):

    # sort the numbers
    # luckily something exists for this...
    numbers.sort()    

    # len(mylist) returns the number of items in it    
    n = len(numbers)  

    # the minimum will be the first item in the list    
    minimum = numbers[0]
   
    # the maximum will be the last item in the list
    maximum = numbers[n - 1]  

    # the median will be the in the n/2 index
    #  notice the integer division!
    median = numbers[n / 2]
	
    # and the quartiles...
    q1 = numbers[n / 4]
    q3 = numbers[3 * n / 4]
	
    # finally, return everything
    return minimum, q1, median, q3, maximum

Lets test our function with a few test cases by running the following instead of main():

print "[min, q1, median, q3, max]"
print calculateStatistics([1, 2, 3, 4, 5, 6, 7, 8, 9])
print calculateStatistics([9, 7, 5, 3, 1, 2, 4, 6, 8])
print calculateStatistics([5])
print calculateStatistics([1, 2])

We get the following as output:

[min, q1, median, q3, max]
(1, 3, 5, 7, 9)
(1, 3, 5, 7, 9)
(5, 5, 5, 5, 5)
(1, 1, 2, 2, 2)

Success!

Example: Box plot

Lets take a step back and see what we have so far.
Notice that we haven't changed any of main or the other functions.
Phew! That's a lot of code... and a decent amount of testing!
Top-down design makes testing easier, because you can test parts.

Most applications in the "real-world" are a hundred, to a thousand, or more times the size of our problem here!

# Filename: boxplot.py
# Author:   Don Miner & Sue Evans
# Date:     8/22/09
# Section:  All
# Email:    bogar@cs.umbc.edu
#
# This program will produce a box plot for the 
# numerical data the user provides either from 
# a file or from typing at the terminal.


# printGreeting() prints an explanation of the program
# Inputs: none
# Output: none
def printGreeting():
    print "\nThis program will produce a box plot"
    print "for the numerical data you enter."
    print "Welcome to my box plot program!\n"


# getInputFromUser() prompts the user for how he will 
# give input, either from a file or from the terminal
# gets the input and returns it.
# Inputs: none
# Output: a list of numbers
def getInputFromUser():

    choice = 'invalid'
    while choice != 'file' and choice != 'type':
       choice = raw_input("Where is the input? (type 'file' or 'type') ")
   
    if choice == 'file':
        return getInputFromFile()
    elif choice == 'type':
        return getInputFromTerminal()


# getInputFromFile() opens the file indicated by the
# user, reads the numerical data from it, closes the
# file and returns the list of numbers.
# Inputs: none
# Output: a list of numbers
def getInputFromFile():
    print "In getInputFromFile()"


# getInputFromTerminal() gets the numerical 
# data from the keyboard as the user enters
# it and returns the list of numbers.
# Inputs: none
# Output: a list of numbers
def getInputFromTerminal():

   n = int(input("How many numbers are there? "))
   
   # initialize an empty list
   numberList = []
   
   for i in range(n):
      newNumber = int(input('> '))
      
      # add this new number to the end of the number list
      numberList.append(newNumber)

   # return the number list
   return numberList


# calculateStatistics() takes a list of numbers and
# calculates the minimum, the first quartile, the
# median, the third quartile and the maximum
# Inputs: a list of numbers called numbers
# Outputs: minimum, q1, median, q3, maximum
def calculateStatistics(numbers):

    # sort the numbers
    # luckily something exists for this...
    numbers.sort()    

    # len(mylist) returns the number of items in it    
    n = len(numbers)  

    # the minimum will be the first item in the list    
    minimum = numbers[0]
   
    # the maximum will be the last item in the list
    maximum = numbers[n - 1]  

    # the median will be the in the n/2 index
    #  notice the integer division!
    median = numbers[n / 2]
	
    # and the quartiles...
    q1 = numbers[n / 4]
    q3 = numbers[3 * n / 4]
	
    # finally, return everything
    return minimum, q1, median, q3, maximum


# printBoxPlot() takes the 5 statistical values, 
# prints them and the box plot for them.
# Inputs: minimum, q1, median, q3, maximum
# Output: none
def printBoxPlot(minimum, q1, median, q3, maximum):

    printBoxPlotGraph(minimum, q1, median, q3, maximum)

    print "min:", minimum
    print "Q1:", q1
    print "median:", median
    print "Q3:", q3
    print "max:", maximum


# printBoxPlotGraph() takes the 5 statistical values
# and prints just the box plot graph for them.
# Inputs: minimum, q1, median, q3, maximum
# Output: none
def printBoxPlotGraph(minimum, q1, median, q3, maximum):

    # find the sizes of the pieces
    whiskerLeftSize = q1 - minimum
    boxLeftSize = median - q1
    boxRightSize = q3 - median
    whiskerRightSize = maximum - q3
   
    # construct the pieces
    # the space before the minimum
    blankLeft = ' ' * minimum 
    whiskerLeft = '-' * whiskerLeftSize
    boxLeft = ' ' * boxLeftSize
    boxRight = ' ' * boxRightSize
    whiskerRight = '-' * whiskerRightSize
   
    # concatenate the pieces
    # break this into left & right so lines don't wrap
    left = blankLeft + '|' + whiskerLeft + '[' + boxLeft + '|'
    right = boxRight + ']' + whiskerRight + '|'
    boxplot = left + right
   
    print boxplot



def main():

   printGreeting()
	
   numberList = getInputFromUser()

   minimum, q1, median, q3, maximum = calculateStatistics(numberList)

   printBoxPlot(minimum, q1, median, q3, maximum)

   
main()

Example: Box plot (Sample Output)

Here is some sample output for our problem.

linuxserver1.cs.umbc.edu[111] python boxplot.py

This program will produce a box plot
for the numerical data you enter.
Welcome to my box plot program!

Where is the input? (type 'file' or 'type') type
How many numbers are there? 9
> 1
> 2
> 3
> 4
> 5
> 6
> 7
> 8
> 9
 |--[  |  ]--|
min: 1
Q1: 3
median: 5
Q3: 7
max: 9
linuxserver1.cs.umbc.edu[112]

`numbers.dat` - (file input)

Right now the user can enter data at the keyboard, but if the user wants to use the program to analyze a lot of data, it will be easier for him/her to put that data in a file. That way making a typo won't cause the user to have to start the entry all over again.
I used the pseudo-random number generator to produce 100 integers in the range of 1 to 19, inclusive, and have stored them in a file called numbers.dat Here it is:

Example: Box plot `getInputFromFile()`

So we only have one more function to write, getInputFromFile()

When the user enters 'file' as his/her choice, we'll need to :

prompt for a filename
open that file for reading
read in each of the numbers in the file
append each new number to the list
close the file
return the list of numbers

# getInputFromFile() opens the file indicated by the
# user, reads the numerical data from it, closes the
# file and returns the list of numbers.
# Inputs: none
# Output: a list of numbers
def getInputFromFile():

    # initialize an empty list
    numberList = []
    
    filename = raw_input("Enter the filename : ")

    file = open(filename, "r")
    for number in file:
        newNumber = eval(number)
        numberList.append(newNumber)

    file.close()
    
    return numberList

As before, we can test this function independently from the rest of the code by calling the function instead of main()


list = getInputFromFile()
print list

linuxserver1.cs.umbc.edu[135] python boxplot.py
Enter the filename : numbers.dat
[2, 17, 2, 12, 16, 8, 18, 12, 18, 8, 14, 11, 17, 18, 13, 4, 14, 13, 12, 11, 10, 6, 6, 7, 19, 13, 9, 17, 9, 1, 17, 2, 11, 9, 8, 6, 11, 12, 17, 17, 19, 3, 14, 5, 19, 9, 17, 7, 11, 15, 8, 9, 16, 6, 17, 2, 6, 11, 8, 16, 3, 11, 14, 9, 3, 10, 13, 6, 10, 3, 2, 10, 8, 15, 11, 19, 5, 13, 6, 7, 6, 7, 2, 13, 1, 10, 13, 9, 15, 15, 17, 15, 3, 13, 4, 18, 14, 17, 8, 19]
linuxserver1.cs.umbc.edu[136]

Success!

Example: Box plot `boxplot.py`

So, we started by writing main() and stubs for the four big parts of our program. main() should always contain the first level of the design which is a skeleton of what the entire program will do.
We chose a part that we wanted to work on first. You don't have to work on the parts in order. I often choose to work on the hardest part first, while I'm fresh and alert.
We chose to work on printBoxPlot() first and found that figuring out how to draw the box plot graph and then drawing it was complicated enough that it should be in its own function. So we identified a subsubproblem, printBoxPlotGraph(), and at first just made a stub for it while we finished writing printBoxPlot().
Working on one function at a time is known as incremental programming. If you adequately test each function when you've finished writing it, you won't have to make any changes. You know it's fine and finished. So any errors that occur will be found in the function you're currently working on, not in others that you've already completed. This certainly helps narrow down the lines of code you need to look at when trying to resolve an error.
Now that we have discovered all of the subproblems and subsubproblems, what does the design of this program look like ?

Testing

Even though we have tested each of the functions separately, we should now test the program as a whole. Let's change the call at the bottom of our code to call main() now and run it.

Here's the output:

ite207-pc-01.cs.umbc.edu[136] python boxplot.py

This program will produce a box plot
for the numerical data you enter.
Welcome to my box plot program!

Where is the input? (type 'file' or 'type') type
How many numbers are there? 9
> 1
> 2
> 3
> 4
> 5
> 6
> 7
> 8
> 9
 |--[  |  ]--|
min: 1
Q1: 3
median: 5
Q3: 7
max: 9
ite207-pc-01.cs.umbc.edu[137] python boxplot.py

This program will produce a box plot
for the numerical data you enter.
Welcome to my box plot program!

Where is the input? (type 'file' or 'type') file
Enter the filename : numbers.dat
 |------[    |    ]----|
min: 1
Q1: 7
median: 11
Q3: 15
max: 19
ite207-pc-01.cs.umbc.edu[138]

Without top-down design...

What would this code look like without top-down design?
How would you change something? Where does one part begin and where does one part end?
What if we wanted to reuse a particlar part of code?
What if we had a bug -- how would you find out where the problem occurred ?

def main():
    
    print "\nThis program will produce a box plot"
    print "for the numerical data you enter."
    print "Hello, welcome to my box plot program!\n"
	
    choice = 'invalid'
    while choice != 'file' and choice != 'type':
        choice = raw_input("Where is the input? (type 'file' or 'type') ")

    numbers = []
    if choice == 'file':
        filename = raw_input("Enter the filename : ")

        file = open(filename, "r")
        for number in file:
            newNumber = eval(number)
            numbers.append(newNumber)

        file.close()
    elif choice == 'type':
        n = int(input("How many numbers are there? "))
     
        for i in range(n):
            newNumber = int(input('> '))
      
            # add this new number to the end of the number list
            numbers.append(newNumber)

    # sort the numbers
    # luckily something exists for this...
    numbers.sort()    

    # len(mylist) returns the number of items in it
    n = len(numbers)  
                  

    # the minimum will be the first item in the list    
    minimum = numbers[0]

    # the maximum will be the last item in the list
    maximum = numbers[n - 1]  

    # the median will be the in the n/2 index
    #  notice the integer division!
    median = numbers[n / 2]

    # and the quartiles...
    q1 = numbers[n / 4]
    q3 = numbers[3 * n / 4]

    # find the sizes of the pieces
    whiskerLeftSize = q1 - minimum
    boxLeftSize = median - q1
    boxRightSize = q3 - median
    whiskerRightSize = maximum - q3

    # construct the pieces
    # the space before the minimum
    blankLeft = ' ' * minimum 
    whiskerLeft = '-' * whiskerLeftSize
    boxLeft = ' ' * boxLeftSize
    boxRight = ' ' * boxRightSize
    whiskerRight = '-' * whiskerRightSize

    # concatenate the pieces
    left = blankLeft + '|' + whiskerLeft + '[' + boxLeft + '|' 
    right = boxRight + ']' + whiskerRight + '|'
    boxplot = left + right

    print boxplot

    print "min:", minimum
    print "Q1:", q1
    print "median:", median
    print "Q3:", q3
    print "max:", maximum

Design Exercise

Draw a design diagram for the following project:

Project description:

This project allows the user to compute the volume and surface area of boxes, when s/he inputs the length, width and height of a box. The user can continue to enter the dimensions of new boxes until s/he is done.

Random Numbers

In our top-down design exercise, we used a file that contained 100 randomly-generated numbers.
A pseudorandom number generator starts with a seed value. It is passed to a function that produces a pseudorandom number. The next time a random number is needed the current value is passed to the function to produce a new random number, similar to the chaos.py program you traced for your first homework.
Games and simulations often require using random numbers.
Python has a random module with some useful functions.
- random.random() which returns a floating point number in the range [0.0, 1.0). This syntax means it includes 0.0, and goes up to, but does not include 1.0

# Filename: rand1.py
# Author:   Sue Evans
# Date:     10/20/09
# Section:  All
# Email:    bogar@cs.umbc.edu
#
# This program illustrates python automatically seeding
# the random number generator using the system's time,
# so the seed will be different everytime the program runs
# using random()

from random import random

for i in range(10):
    for j in range(5):

        number = random()
        print "%.12f" % (number),
        
    print

Here's the output from two runs:

linuxserver1.cs.umbc.edu[128] python rand1.py
0.943261480372 0.154005266066 0.844525974976 0.583765951220 0.238710918762
0.840630709372 0.792588202641 0.922412700010 0.406105887829 0.928238267395
0.368546449511 0.436283507343 0.134233342198 0.241690027160 0.549987196778
0.680875157308 0.891671479328 0.163688215656 0.356264886504 0.917510472125
0.536991538634 0.725870958470 0.061311983526 0.467390124198 0.841849165201
0.736892870516 0.616416745251 0.765799219193 0.688528381762 0.335759874218
0.713517554627 0.873412461417 0.946687666341 0.753602080884 0.797920356699
0.944693652515 0.065424134275 0.024200402342 0.708441209872 0.927037309789
0.455257650896 0.222231393534 0.920537903693 0.491643880902 0.244733255432
0.284381032317 0.524329119253 0.372705775886 0.079492870458 0.134190063063
linuxserver1.cs.umbc.edu[129] python rand1.py
0.822015872381 0.236398450478 0.496277031584 0.652754711980 0.430250909378
0.712871191620 0.456168190880 0.852566313990 0.643437647359 0.437319556849
0.193554550051 0.970620848238 0.008477434438 0.550374015106 0.429892280832
0.367912143435 0.263518888582 0.690646019624 0.301095452514 0.799951784722
0.553873714449 0.756633155350 0.044985173260 0.137729652020 0.880460160921
0.749262682891 0.620612164245 0.057671190146 0.494361182914 0.471051448670
0.217800777087 0.950471608505 0.086212696549 0.702363377091 0.310036957841
0.065677666387 0.661299499973 0.425212880176 0.499324519028 0.269682940655
0.366411218288 0.093996635968 0.180165188941 0.626948514247 0.688864919803
0.979156243757 0.905714196934 0.191048791602 0.601074564485 0.598188508617
linuxserver1.cs.umbc.edu[130]

Here's an example that uses random.randrange([start], stop[, step] where start and step are optional. The default start is 0 and step is 1, just as with range().

# Filename: rand2.py
# Author:   Sue Evans
# Date:     10/20/09
# Section:  All
# Email:    bogar@cs.umbc.edu
#
# This program illustrates python automatically seeding
# the random number generator using the system's time,
# so the seed will be different everytime the program runs
# using randrange() with a range of 1 to 1000, inclusive.

from random import randrange

for i in range(10):
    for j in range(5):

        number = randrange(1, 1001)
        print "%7d" % (number),
        
    print

Here's the output :

linuxserver1.cs.umbc.edu[133] python rand2.py
    688       7     220     438     586
    976     980     529     656     895
    241     236     638     593     116
    749     432     668     841     177
    345      51     388     590      16
     61     665     351     797     897
    183     841     389     845     135
    537     132     298     847     997
    675     196     288     823     428
    254     770     967     952      97
linuxserver1.cs.umbc.edu[134] python rand2.py
    393     504     112      67     306
    917     998     144     576     972
    546     975     978     540     194
    764     269     196     187     835
     62     396     544     469     779
    545     489     264     609     434
    667     124     850      35     830
    778      56     713     163     753
    276     413     381     271     254
    649     874     207     443     406
linuxserver1.cs.umbc.edu[135]

Random Exercise

Write python code that will write 100 randomly-generated values between 1 and 19, inclusive, to a file called numbers.dat with one number per line.

Top-Down Design

Learning Objectives

Top-Down Design

Top-Down Design

Example problem: Box plot

Example: Box plot (specification)

Example: Box plot (first level)

Example: Box plot (printBoxPlot())

Example: Box plot (printBoxPlotGraph())

Example: Box plot

Example: Box plot (getInputFromUser())

Example: Box plot (getInputFromTerminal())

Example: Box plot (calculateStatistics())

Example: Box plot

Example: Box plot (Sample Output)

numbers.dat - (file input)

Example: Box plot getInputFromFile()

Example: Box plot boxplot.py

Testing

Without top-down design...

Design Exercise

Random Numbers

Random Exercise

Example: Box plot (`printBoxPlot()`)

Example: Box plot (`printBoxPlotGraph()`)

Example: Box plot (`getInputFromUser()`)

Example: Box plot (`getInputFromTerminal()`)

Example: Box plot (`calculateStatistics()`)

`numbers.dat` - (file input)

Example: Box plot `getInputFromFile()`

Example: Box plot `boxplot.py`