Question

Problem 1: Implement count-min sketch Implement CountMinSketch class below where num_counters is the number of counters. You are given the constructor that already generates a random representative of a hash function family. Implement the functions: * increment * approximateCount Please read the constructor carefully: it initializes the counters and generates the hash function for you. Also, when you call hash_string function defined previously, do not forget to take result modulo m. # Class for implementing a count min sketch "single bank" of counters class CountMinSketch: # Initialize with `num_counters` def init(self, num_counters): self.m = num_counters self.hash_fun_rep = get_random_hash_function() self.counters = [0]*self.m # your code here # function: increment # given a word, increment its count in the countmin sketch def increment(self, word): # your code here # function: approximateCount # Given a word, get its approximate count def approximateCount(self, word): # your code here # We will now implement the algorithm for a bank of k counters # Initialize k different counters def initialize_k_counters(k, m): return [CountMinSketch(m) for i in range(k)] # Function increment_counters # increment each of the individual counters with the word def increment_counters(count_min_sketches, word): # your code here # Function: approximate_count # Get the approximate count by querying each counter bank and taking the minimum def approximate_count(count_min_sketches, word): return min([cms.approximateCount(word) for cms in count_min_sketches]) %matplotlib inline from matplotlib import pyplot as plt # Let's see how well your solution performs for the Great Gatsby words cms_list = initialize_k_counters(5, 1000) for word in longer_words_gg: increment_counters(cms_list, word) discrepancies = [] for word in longer_words_gg: l = approximate_count(cms_list, word) r = word_freq_gg[word] assert ( l >= r) discrepancies.append( l-r ) plt.hist(discrepancies) assert(max(discrepancies) <= 200), 'The largest discrepancy must be <= 200' print('Passed all tests: 10 points')

          Problem 1: Implement count-min sketch

Implement CountMinSketch class below where num_counters is the number of counters. You are given the constructor that already generates a random representative of a hash function family. Implement the functions:

* increment
* approximateCount

Please read the constructor carefully: it initializes the counters and generates the hash function for you. Also, when you call hash_string function defined previously, do not forget to take result modulo m.

# Class for implementing a count min sketch "single bank" of counters
class CountMinSketch:
    # Initialize with `num_counters`
    def __init__(self, num_counters):
        self.m = num_counters
        self.hash_fun_rep = get_random_hash_function()
        self.counters = [0]*self.m

    # your code here

    # function: increment
    # given a word, increment its count in the countmin sketch
    def increment(self, word):
        # your code here

    # function: approximateCount
    # Given a word, get its approximate count
    def approximateCount(self, word):
        # your code here

# We will now implement the algorithm for a bank of k counters

# Initialize k different counters
def initialize_k_counters(k, m):
    return [CountMinSketch(m) for i in range(k)]

# Function increment_counters
# increment each of the individual counters with the word
def increment_counters(count_min_sketches, word):
    # your code here

# Function: approximate_count
# Get the approximate count by querying each counter bank and taking the minimum
def approximate_count(count_min_sketches, word):
    return min([cms.approximateCount(word) for cms in count_min_sketches])

%matplotlib inline
from matplotlib import pyplot as plt

# Let's see how well your solution performs for the Great Gatsby words
cms_list = initialize_k_counters(5, 1000)
for word in longer_words_gg:
    increment_counters(cms_list, word)

discrepancies = []
for word in longer_words_gg:
    l = approximate_count(cms_list, word)
    r = word_freq_gg[word]
    assert ( l >= r)
    discrepancies.append( l-r )

plt.hist(discrepancies)

assert(max(discrepancies) <= 200), 'The largest discrepancy must be <= 200'
print('Passed all tests: 10 points')

Problem 1: Implement count-min sketch

Implement CountMinSketch class below where numcounters is the number of counters. You are given the constructor that already generates a random representative of a hash function family. Implement the functions:

* increment
* approximateCount

Please read the constructor carefully: it initializes the counters and generates the hash function for you. Also, when you call hashstring function defined previously, do not forget to take result modulo m.

# Class for implementing a count min sketch "single bank" of counters
class CountMinSketch:
# Initialize with `numcounters`
def init(self, numcounters):
self.m = numcounters
self.hashfunrep = getrandomhashfunction()
self.counters = [0]*self.m

# your code here

# function: increment
# given a word, increment its count in the countmin sketch
def increment(self, word):
# your code here

# function: approximateCount
# Given a word, get its approximate count
def approximateCount(self, word):
# your code here

# We will now implement the algorithm for a bank of k counters

# Initialize k different counters
def initializekcounters(k, m):
return [CountMinSketch(m) for i in range(k)]

# Function incrementcounters
# increment each of the individual counters with the word
def incrementcounters(countminsketches, word):
# your code here

# Function: approximatecount
# Get the approximate count by querying each counter bank and taking the minimum
def approximatecount(countminsketches, word):
return min([cms.approximateCount(word) for cms in countminsketches])

%matplotlib inline
from matplotlib import pyplot as plt

# Let's see how well your solution performs for the Great Gatsby words
cmslist = initializekcounters(5, 1000)
for word in longerwordsgg:
incrementcounters(cmslist, word)

discrepancies = []
for word in longerwordsgg:
l = approximatecount(cmslist, word)
r = wordfreqgg[word]
assert ( l >= r)
discrepancies.append( l-r )

plt.hist(discrepancies)

assert(max(discrepancies) <= 200), 'The largest discrepancy must be <= 200'
print('Passed all tests: 10 points')

Added by Noelia H.

Question

Please give Ace some feedback