README.md

My Python Practice

I'm sorry R …… I betrayed you.

<List>

• Shuffle List (2020.03.30)
• Random Seed Influence (2020.01.05)
• Square Root (2020.01.01) (adjusted 2020.01.04)
• Limited Range Sampling (2019.09.22)

Shuffle List (2020.03.30)

• find how to get random lists without overlapping values

• use random random.randint random.sample

  SuffleList.py

  import random
  Trial 1 : Use random.randint()

  shufflelist1 = []
  
  for i in range(0,20) :
      random.seed(330 + i)
      shufflelist1.append(random.randint(1, 20))
  
  print(shufflelist1) # There are overlapping values.

  [20, 11, 8, 18, 8, 5, 1, 7, 4, 5, 13, 19, 4, 7, 13, 10, 18, 12, 11, 14]

  Trial 2 : Use random.sample()

  random.seed(330)
  shufflelist2 = random.sample(range(1, 21), 20)
  
  print(shufflelist2) # random.sample() offers values without overlapping.

  [20, 3, 2, 13, 1, 6, 10, 9, 15, 11, 14, 4, 18, 8, 16, 17, 7, 19, 12, 5]

  Trial 3 : Use while Statement

  shufflelist3 = []
  loopnum = 0
  
  while len(shufflelist3) < 20 :
      random.seed(330 + loopnum)
      r = random.randint(1,20)
      if r not in shufflelist3 : shufflelist3.append(r)
      loopnum += 1
  
  print(shufflelist3)
  # It seems similar with Trial 1's sequence but there's no overlapping values.

  [20, 11, 8, 18, 5, 1, 7, 4, 13, 19, 10, 12, 14, 6, 2, 3, 17, 16, 15, 9]

  print(loopnum) # It shows how many times overlapping numbers are rejected.

  87

Random Seed Influence (2020.01.05)

• Make sure the range of random.seed()'s influence
  ☞ random.seed() affects just one time!

  Code : RandomSeedInfluence.py

  import random

  # case 1
  print(random.random())
  print(random.random())
  print(random.random())

  0.48515227527760874
  0.48808537244754757
  0.9509662749522355

  # case 2
  random.seed(105)
  print(random.random())
  print(random.random())
  print(random.random())

  0.8780993490764925
  0.3491186468357038
  0.7907236599059974

  # case 2-1
  random.seed(105); print(random.random())
  random.seed(105); print(random.random())
  random.seed(105); print(random.random())

  0.8780993490764925
  0.8780993490764925
  0.8780993490764925

  # case 3
  random.seed(105)
  for i in range(0,3) :
      print(random.random())

  0.8780993490764925
  0.3491186468357038
  0.7907236599059974

  # case 3-1
  for i in range(0,3) :
      random.seed(105); print(random.random())

  0.8780993490764925
  0.8780993490764925
  0.8780993490764925

Square Root (2020.01.01)

• An algorithm to find n's square root without math.sqrt()

• Adjusted 2020.01.04 : rearrange methods' order in for Loop for improving intuitive understanding

  Code : SquareRoot.py

  import random
  import math
  import matplotlib.pyplot as plt
  
  n = 2 # should be larger than 1
  k = 20 # run loop k times
  
  squareroot = []
  lowerlimit, upperlimit = 1, n
  
  for i in range(k) :
  
      random.seed(20200104) # can be removed
      squareroot.append(random.uniform(lowerlimit, upperlimit))
      square = squareroot[i] ** 2
      print(i+1, squareroot[i], square, square-n)
  
      if square == n :
          break;
      elif square < n :
          # print("smaller")
          lowerlimit = max(squareroot[i], lowerlimit)
      else :
          # print("larger")
          upperlimit = min(squareroot[i], upperlimit)
  
  myplot = plt.plot(range(k), squareroot)
  # myplot.hlines(math.sqrt(n), color="red", linestyle="--") # doesn't work

  1 1.224709461308563 1.4999132646187106 -0.5000867353812894
  2 1.3989245806155413 1.956989982250368 -0.04301001774963198
  3 1.5339919143112415 2.3531311931722674 0.3531311931722674
  (중략)
  19 1.4141854421168503 1.9999204646952313 -7.953530476867421e-05
  20 1.4141980335178153 1.9999560780056558 -4.3921994344220394e-05

  

  Practice

  # practice
  random.random()
  random.randrange(1,n) # output only integer
  random.uniform(1,n) # output float
  list(range(10))

  0.2508550895840985
  1
  1.2710268293926659
  [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]

Limited Range Sampling (2019.09.22)

• Generate normal distributed sample with limited range

• Use numpy matplotlib.pyplot scipy

  LimitedRangeSampling.py

  Generate a normal distribution with limited range [25, 75]

  import numpy as np
  import matplotlib.pyplot as plt
  from scipy import stats
  
  mu, sigma, n = 50, 10, 1000
  llimit, rlimit = 25, 75
  
  data = np.random.normal(mu, sigma, n)
  Method 0. Generating initial data (not trimmed yet)

  plt.hist(data)
  stats.describe(data)[0:2] # [0] : nobs, [1] : minmax

  

  (1000, (16.763171096395133, 76.969552776105601))

  Method 1. Trim with rack of amount

  data1 = data[(data >= llimit) & (data <= rlimit)]

  plt.hist(data1)
  stats.describe(data1)[0:2]

  

  (991, (25.600374595125377, 74.942171158969671))

  Method 2. Check each one trial

  data2, amount = [], 0
  
  while amount < n :
      data_temp = np.random.normal(mu, sigma, 1)
      if (data_temp >= llimit) & (data_temp <= rlimit) :
          data2 = np.append(data2, data_temp)
          amount += 1

  plt.hist(data2)
  stats.describe(data2)[0:2]

  

  (1000, (25.987274047611137, 73.473315070409228))

  Method 3. Generate one round and fill the lack

  data3 = data[(data >= llimit) & (data <= rlimit)]
  amount = len(data3)
  
  while amount < n :
      data_temp = np.random.normal(mu, sigma, 1)
      if (data_temp >= llimit) & (data_temp <= rlimit) :
          data3 = np.append(data3, data_temp)
          amount += 1

  plt.hist(data3)
  stats.describe(data3)[0:2]

  

  (1000, (25.600374595125377, 74.942171158969671))