Noteathon / Python-Crash-Course

Python-Crash-Course

August 19, 2021

Python : Python Crash Course

Initial setup
  • Is using anaconda then try using the below command as it will setup all the initial base required packages 
    conda update conda
conda update anaconda
conda create -n python_udemy_course anaconda

Python _ 1

Python _ 2

  • List []
    • Can support different datatypes, item assignment and deletion are possible
    • append to add.
  • Tuples ()
    • I dont know why they are used.
    • Once assigned stays fixed, can go for any item modification.
    • Hashable since they are immutable, e.g. in key value pair you want to use, which you might never.
  • Dictionary {}
    • simple key:value pair, can be updated and added new keys.
    • update() to update or can directly add d[key] = value
  • Set {}
    • set( iteratable type ) function to get the unique items
    • add(elem) to add
  • Comparision Operators <= > != and or 
    print(True or False)
print(True and False)
print(1<2 and 2>3)
print(1==1)

if False:
  print('Perform some operation')
elif 1 ==1 :
    print('my middle operation')
else:
    print(' You are not True')
  • Loops for, while
    • Use enumerate to get the idx in for loop 
      items = ['apple','bananna','mango','strawberry']
for idx, item in enumerate(items, start=0):
  print(idx,",",item)
 # enumerate object yields pairs containing a count (from start, which defaults to zero) and a value yielded by the iterable argument
enum_obj = enumerate(items) 
var1, var2 = next(enum_obj) # unpacks tuple (0, seq[0])
print(var1, var2)
# List append
out = [i**2 for i in range(10)]
  • Functions, lambdas, maps 
    print(list(map(my_func, seq1)))
## Convert to lambda
print(list(map(lambda var: var**2, seq1)))
## To filter even numbers
print(list(filter(lambda var : var%2==0, seq1)))
## Chek items present in list
item in list # return True or False
## Tuple unpacking
lst_tuple = [(1,3),(5,7),(9,11)]
for i,j in lst_tuple:
    print(i,j)
## Use of ternary operators, if birthday then the bound can + 5
def caught_speeding(speed, is_birthday):
  lowBound = 60 + ( 5 if is_birthday else 0)
  highBound = 80 + ( 5 if is_birthday else 0)
        
  if (speed <= lowBound):
      return "No ticket"
  elif (speed >lowBound and speed <=highBound):
      return "Small Ticket"
  else:
      return "Big Ticket"

    Python _ 3 Numpy

  • Setup 
    # conda install numpy
# pip install numpy
# conda list numpy
# import numpy
# numpy.__version__
  • Flavours
    • Vector : strictly 1 D array
    • Matrix : 2-D but can have 1-D too.
  • Array
    • single list :
    • multi list : 
      np.array([10,20]) # [10 20]
np.array([[10,20], [30,40]]) # array([[10, 20],[30, 40]])
# zero to 11 with step 2
print(np.arange(0,11,2)) # [ 0  2  4  6  8 10]
print(np.zeros(10)) #[0. 0. 0. 0. 0. 0. 0. 0. 0. 0.]
print(np.ones((3,4))) # [[1. 1. 1. 1.] [1. 1. 1. 1.] 1. 1. 1. 1.]]
# Difference btw arange and linspace, in arnage we specify the step size whereas in linspace we specify the number of elems we need in that range.
print(np.linspace(0, 5, 10)) # [0.         0.55555556 1.11111111 1.66666667 2.22222222 2.77777778 3.33333333 3.88888889 4.44444444 5.        ]
print(np.eye(4)) # identity matrix [[1. 0. 0. 0.] [0. 1. 0. 0.] [0. 0. 1. 0.] [0. 0. 0. 1.]]

print(np.random.rand(5))
print(np.random.rand(2,3)) # get 5 random nos [0.26320601 0.98699869 0.03921547 0.59362448 0.47860516]
print(np.random.randn(6)) # normal distribution points
# print rand int between range
print(np.random.randint(0,4)) # can be 0,1,2,3
# print 5 rand int between range
print(np.random.randint(0,10,5)) # [2 5 2 2 5]
# print 5 rand int between range
arr = np.random.randint(0,20,10) # [2 5 2 2 5]
print(arr)
print(arr.reshape(2,5))

"""
[13  0  9  6  3  6  6  5  3  4]
[[13  0  9  6  3]
[ 6  6  5  3  4]]
"""
np.array(lst).min()
np.array(lst).max()
## To know the idx
np.array(lst).argmax()
np.array(lst).argmin()
# Shape to know the dims (rows, cols)
reanr.shape # (2,5)
# To know the datatype
reanr.dtype # dtype('int64')

# To make a new copy of the array and use separately, use .copy else only reference will be changed.
arr1 = arr.copy()
# Accessing array elements
print(arr[0,2]) or print(arr[0][2])
# To access all the rows from 0 - 2 and from column 1 
print(arr[:2,1:])
#  [[ 0  1  2  3  4  5  6  7  8  9]
# [10 11 12 13 14 15 16 17 18 19]
# [20 21 22 23 24 25 26 27 28 29]
# [30 31 32 33 34 35 36 37 38 39]
# [40 41 42 43 44 45 46 47 48 49]]
# In the above arr to grab the array [[24 25] [34 35]]
arr_2d = np.arange(50).reshape(5,10)

print(arr_2d)
print("")
print(arr_2d[2:4,4:6])
# To print the last column
print(arr_2d[:,-1:])
    
# To do the array boolean operations
arr = np.arange(0,10)
print(arr)
# [0 1 2 3 4 5 6 7 8 9]
arr = arr[arr > 5]
print(arr)
# [6 7 8 9]
# To get the sum of all elements in the column
arr_2d.sum(axis=0)
    • Operations
      • array with array 
        arr = np.arange(0,10)
print(arr)
# [0 1 2 3 4 5 6 7 8 9]
print(arr + arr)
# [ 0  2  4  6  8 10 12 14 16 18]
print(arr * arr)
# [ 0  1  4  9 16 25 36 49 64 81]
print(arr / arr) # WARNING invalid value encountered in true_divide which give nan
# [nan  1.  1.  1.  1.  1.  1.  1.  1.  1.]
      • scalar operations 
        print(arr ** 3)
# [  0   1   8  27  64 125 216 343 512 729]
      • Universal array functions You can find the link, which has all the various operations involved such as Math, Trignometric, Bitwise
      print(np.sqrt(arr))
# [0.         1.         1.41421356 1.73205081]
print(np.max(arr))
# 3

      Python _ 4 Pandas

  • Setup 
    # conda install pandas
# pip install pandas
  • Series 
    import numpy as np
import pandas as pd
labels = ['a','b','c']
my_data = np.arange(0,3)
ser1 = pd.Series(my_data, labels)
print(ser1)
# a    0
# b    1
# c    2
# dtype: int64
ser1 = pd.Series([1,2,3,4],index = ['USA', 'Germany','USSR', 'Japan'])                                   
ser2 = pd.Series([1,2,5,4],index = ['USA', 'Germany','Italy', 'Japan'])
print(ser1+ser2)
# In the below if any value is not present, then will returnn NaN for that operation.
# Germany    4.0
# Italy      NaN
# Japan      8.0
# USA        2.0
# USSR       NaN
# dtype: float64

Useful commands

# To list envs in conda
conda info --envs

# To install pytorch
conda install pytorch torchvision torchaudio -c pytorch

References

Next: Algorithms