Parallel computing hacks (start here!)

Hacks

AP Classroom. Provide answers and thoughts on theoritical question form college board Video in section 4.3. They start at about the 9 minute mark.

  • Example 1
  • Example 2

Data Structures. Build a List Comprehension example

  • list = [calc(item) for item in items]

HACCCS response

notes from video

Sequential computing is a computational model in which operations are performed in order one at a time

Parallel computing is a computational model where the program is broken into multiple smaller sequential computing operations,, some of which are performed simultaneously.

Computers handle a lot of tasks at once

running tasks can be done sequentially, in parallel.

tasks are done one at a time in sequential programming (like conveyor belt)

IN PARALLEL COMPUTING:parallel computing schedules tasks to be executed at same tme parallel computing overall processes data faster and it scales more effectively.

simple Example

Core 1: task A takes 45 ms

Core 2: task B takes 50 ms

Core 3: Task C takes 25 ms; task D takes 30 ms

Core 4: Task E takes 40 ms

Total time: 55 ms (core 3 runs task C and D in sequence)

Basically, 5 tasks are spread out between 4 cores, and all complete tasks at once, finally being done when all cores finish their individual task.

Obviously, if we run all tasks in sequential on 1 core, the time becomes much longer.

Distributive computing: we now distribute our tasks among dfferent computers, not just among 1 physical computer. This needs network connection. Can take longer if network is bad.

Example 1

A particular computer has 2 identical processors which can run in parallel. Each process must be executed on 1 processor and each processor can only run 1 process at a time.

whats the minimum amount of time (approx) to execute all 3 processes when 2 processor run in paallel?

process X: 50 sec

process Y: 10 sec

process Z: 30 sec

I think this would take 50 seconds, as we can run process X on one processor while the other runs Y and Z.

I WAS RIGHT!

This is a pretty good example for why parallel is good, but how can we develop programs for parallel processing if we don't have multiple processors to test on?

Example 2

What is the difference in execution time between runniing the 2 processes in parallel in place of running on single processor?

sequential: 70 seconds

Parallel: 45 seconds

70 - 45 = 25 seconds

My ans: difference of 25 seconds

I was right

difference of 25 seconds is pretty large and seems unrealistic, but I think it demonstrates why we need to optimize for parallel computing.

list comprehension example

Data Structures. Build a List Comprehension example

  • list = [calc(item) for item in items]
items = ["cinnamon", "cupcake batter", "fig"]

numbers = [1,2,20,40]

newlist = [x  if x != "fig" else "date" for x in items] # remove fig from items

numlist = [x for x in numbers]

sum = 0
for y in numbers:
    sum = sum + y
    print(sum)


print(newlist)

1
3
23
63
['cinnamon', 'cupcake batter', 'date']

END OF PARALLEL HACKS

START OF HASHMAP HACKS

Hacks

  • Answer ALL questions in the code segments
  • Create a diagram or comparison illustration (Canva).
    • What are the pro and cons of using this data structure?
    • Dictionary vs List
  • Expand upon the code given to you, possible improvements in comments

  • Build your own album showing features of a python dictionary

  • For Mr. Yeung's class: Justify your favorite Taylor Swift song, answer may effect seed

working on hacks

img

creating own album with python 

my_album = { # initialize the album need to run each time
    "title": "I know how to use dictionaries",
    "artist": "artist",
    "year": 1984,
    "genre": [],
    "tracks": {
        1: "U.N Owen Was her?",
    }
}

def search_personal():
    x = 1
    while x == 1: # use while loop so we can continue searching
        search = input("What would you like to know about the album? (type quit to leave)")
        if search == "quit":
            print("you have quit lol")
            x = 0
        elif my_album.get(search.lower()) == None:
            print("Invalid Search")
        else:
            print(my_album.get(search.lower()))
    
       
#search_personal()

my_tracks = my_album["tracks"]
def addtracks():
    x = 1
    while x == 1: # use while loop so we can continue searching
        newTrack = input("add a new track (type tracks), change artist (type artist), show album (type search), add genre (type genre) or type quit")
        
        if newTrack == "quit":
            x = 0
        
        elif newTrack == "artist":
            artists = input("type in artist name")
            my_album["artist"] = artists
        
        elif newTrack == "tracks":
            newTrackz = input("type in new track name")
            newTrack2 = input("add position of new track:")
            my_tracks[newTrack2] = newTrackz
        
        elif newTrack == "search":
            search_personal()

        
        elif newTrack == "genre":
            genres = input("add a genre to the program")
            my_album["genre"].append(genres)
            newGenre = []
            [newGenre.append(x) for x in my_album["genre"] if x not in newGenre] # use list comprehension to remove duplicates (wow im so good at coding)
            my_album["genre"] = newGenre # add genre
            print(my_album["genre"])
        
        else:
            print("unrecognized command!")

addtracks()
 

#search_personal()
print(my_album["tracks"])

{1: 'U.N Owen Was her?', '2': 'discord', '3': 'maybe'}
Invalid Search
unknown
unknown
{1: 'U.N Owen Was her?', '2': 'discord', '3': 'maybe'}
you have quit lol
['touhou']
['touhou']
{1: 'U.N Owen Was her?', '2': 'discord', '3': 'maybe', '4': 'tomboyish girl in love'}
you have quit lol
{1: 'U.N Owen Was her?', '2': 'discord', '3': 'maybe', '4': 'tomboyish girl in love'}

answering questions 

my_set = set([1, 2, 3, 2, 1])
print(my_set)  

# What do you notice in the output?
# THE output of the function is {1,2,3}. 
# There are no duplicate values 
# probably b/c we can't have duplciates in hashes.

# Why do you think Sets are in the same tech talk as Hashmaps/Hashtables?
# Sets are a good example of how collisions in hashing works and why we shouldn't have duplicate
# Sets are a tool we can use for hashing if we don't want to use dictionaries.

# Creating a dictionary with information about the album "Lover"
lover_album = {
    "title": "Lover",
    "artist": "Taylor Swift",
    "year": 2019,
    "genre": ["Pop", "Synth-pop"],
    "tracks": {
        1: "I Forgot That You Existed",
        2: "Cruel Summer",
        3: "Lover",
        4: "The Man",
        5: "The Archer",
        6: "I Think He Knows",
        7: "Miss Americana & The Heartbreak Prince",
        8: "Paper Rings",
        9: "Cornelia Street",
        10: "Death By A Thousand Cuts",
        11: "London Boy",
        12: "Soon You'll Get Better (feat. Dixie Chicks)",
        13: "False God",
        14: "You Need To Calm Down",
        15: "Afterglow",
        16: "Me! (feat. Brendon Urie of Panic! At The Disco)",
        17: "It's Nice To Have A Friend",
        18: "Daylight"
    }
}

# What data structures do you see?
# Lists
# all of this is compiled into a dictionary

# Printing the dictionary
print(lover_album)

# Retrieve value from dictionary with key
print(lover_album.get('tracks'))
# or
print(lover_album['tracks'])

# adding a value with a new key
lover_album["producer"] = ['Taylor Swift', 'Jack Antonoff', 'Joel Little', 'Taylor Swift', 'Louis Bell', 'Frank Dukes']
# What can you change to make sure there are no duplicate producers?
# Just remove the duplicate manually lol
# or use sets (that worked)

settest = set(lover_album["producer"])


# Printing the dictionary
print(lover_album["producer"])
print(settest)

# Adding a an key-value pair to an existing key 
lover_album["tracks"].update({19: "All Of The Girls You Loved Before"})

# How would add an additional genre to the dictionary, like electropop? 
# 
lover_album["genre"] =["Pop", "Synth-pop", "electropop"] # will work, but maybe there is another way
# 
lover_album["genre"].append("new genre") #this method uses list methods but we now have duplicates if we run multiple times

# Printing the dictionary
print(lover_album["genre"])

# Print lover_album in more readable format
for k,v in lover_album.items(): # iterate using a for loop for key and value
    #print(str(k) + ": " + str(v)) dont need this....
    pass

# Write your own code to print tracks in readable format
# 
#
exe = 0
trackS = lover_album["tracks"]
num = 0

for k,v in trackS.items(): # iterate using a for loop for key and value
    exe = exe + 1
    num = num + 1
    print("song", exe, ":", trackS[num])

# Using conditionals to retrieve a random song
def search():
    x = 1
    while x == 1: # use while loop so we can continue searching
        search = input("What would you like to know about the album? (type quit to exit)")
        if search == "quit":
            x = 0
        elif lover_album.get(search.lower()) == None:
            print("Invalid Search")
        
        else:
            print(lover_album.get(search.lower()))
    
search()

# This is a very basic code segment, how can you improve upon this code?
# use while loop so that we can continue searching
#