7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

Milind Mali / Data Scientist / Data Analyst

Basic operation with NumPy

Numpy- Numerical Python

1.Built in function in numpy array

2.Creating one dimensional array
3.Creating two dimensional array
4.Three dimensional array
5.Higher dimensional array
6.Initial placeholder in numpy array
7.Analysis numpy array
8.Mathematical operation with numpy
9.Array manipulation
10.Numpy Indexing and selection
11.Grabbing single element
12.Broadcasting
13.Slicing
14.Indexing and selection in two dimensional array
15.Indexing in 3D array
16.Condition and selection

Advantage of Numpy Array

Allow several mathematical Operation

faster operations

although array list and tuple all are used to hold the any data but the process time takes for array operation is
comparatively very less

List Vs Numpy Array--> Lets check the time taken

In [2]:

import numpy as np

In [3]:

from time import process_time

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 1/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [26]:

# time taken by list

list1=[i for i in range(10000000)]

start_time=process_time()

list1=[i+10 for i in list1] #adding five to each number

end_time=process_time()

print(end_time-start_time)

1.578125

In [25]:

array1=np.array([i for i in range(10000000)])

start_time=process_time()

array1+=10

end_time=process_time()

print(end_time-start_time)

0.015625

Conclusion :--> required time for operation with array is very less thats why we prefer to array against list or
tuple

In [31]:

list1=[1,2,3,4,5]

print(list1)

print(type(list1))

[1, 2, 3, 4, 5]
<class 'list'>

numeber are seperated by comma

In [32]:

Array1=np.array([1,2,3,4,5])

print(Array1)

print(type(Array1))

[1 2 3 4 5]
<class 'numpy.ndarray'>

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 2/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

number of are not seperated by comma

Built in function in Numpy Array

creating one dimentional array

In [93]:

# arange function
x=np.arange(1,10,2) # from 1 to 10 with step size 2
print(x)

[1 3 5 7 9]

In [94]:

# five random values from zero to one

x=np.random.rand(5)
print(x)

[0.28285425 0.90522102 0.13056326 0.30841338 0.10878923]

this is two dimentional array

In [99]:

list1=[[1,2,3],[4,5,6],[7,8,9]]

my_matrix=np.array(list1)

print(my_matrix)

[[1 2 3]
[4 5 6]
[7 8 9]]

In [95]:

# 2*3 matrix contain from zero to one

x=np.random.rand(2,3)
print(x)

[[0.03096923 0.81105207 0.32229728]

[0.44511911 0.17732007 0.52159635]]

In [96]:

# random values follows std normal distribution means whose mean is zero and std dev is 1
x=np.random.randn(5)
print(x)

[-0.74418913 0.43635588 -1.75289267 0.05025689 1.4464433 ]

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 3/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [97]:

# random values follows std normal distribution means whose mean is zero and std dev is 1
x=np.random.randn(3,4)
print(x)

[[-0.35284373 1.41559546 1.35191402 0.15987834]

[ 0.82225866 -1.14712803 -0.50553481 0.75719029]
[ 0.368709 -0.45620814 -0.20477501 0.4112788 ]]

In [98]:

#np.random.randint
x=np.random.randint(10,100,(3,4))
print(x)

[[93 25 84 86]
[57 74 44 25]
[97 46 47 29]]

In [40]:

# now create two dimentional array (with defining values at float values)
array2=np.array([(1,2,3,4),(5,6,7,8)],dtype=float)

print(array2)

array2.shape

[[1. 2. 3. 4.]
[5. 6. 7. 8.]]

Out[40]:

(2, 4)

Three dimentional array

In [155]:

arr = np.array([[[1, 2, 3], [4, 5, 6]], [[1, 2, 3], [4, 5, 6]]])

print(arr)

[[[1 2 3]
[4 5 6]]

[[1 2 3]
[4 5 6]]]

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 4/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

higher Dimentional arrays

In [160]:

arr = np.array([1, 2, 3, 4], ndmin=5)

print(arr)
print('number of dimensions :', arr.ndim)

[[[[[1 2 3 4]]]]]
number of dimensions : 5

In this array the innermost dimension (5th dim) has 4 elements, the 4th dim has 1 element that is the vector,
the 3rd dim has 1 element that is the matrix with the vector, the 2nd dim has 1 element that is 3D array and
1st dim has 1 element that is a 4D array.

intial placeholder in numpy array

in some cases we want to intiate array with certain values

In [44]:

# create numpy array with all the values are zero

x=np.zeros((4,5))
print(x)

[[0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0.]
[0. 0. 0. 0. 0.]]

In [45]:

# create numpy array with all the values are one

x=np.ones((4,5))
print(x)

[[1. 1. 1. 1. 1.]
[1. 1. 1. 1. 1.]
[1. 1. 1. 1. 1.]
[1. 1. 1. 1. 1.]]

In [46]:

# array of perticular values

x=np.full((4,4),5)
print(x)

[[5 5 5 5]
[5 5 5 5]
[5 5 5 5]
[5 5 5 5]]

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 5/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [48]:

# creating identity matrix- all the diogonal element will be one and rest all the values

x=np.eye(4)
print(x)

[[1. 0. 0. 0.]
[0. 1. 0. 0.]
[0. 0. 1. 0.]
[0. 0. 0. 1.]]

In [50]:

# create array with random values

x=np.random.random((3,4))
print(x)

[[0.73860796 0.10178962 0.98704855 0.23459453]

[0.55524451 0.21111667 0.68869849 0.37894866]
[0.61243223 0.1972518 0.38222988 0.44870634]]

all the values you will get here will be in the range of 0 to 1

In [52]:

# create random int values

x=np.random.randint(10,100,(3,4))
print(x)

[[37 47 35 38]
[13 31 94 24]
[75 93 49 38]]

In [54]:

# creating array of evenly spaced values

x=np.linspace(10,30,5)
print(x)

[10. 15. 20. 25. 30.]

In [55]:

# converting list to array

list1=[10,20,30,40,50]

array1=np.asarray(list1)

print(array1)
type(array1)

[10 20 30 40 50]

Out[55]:

numpy.ndarray

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 6/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [56]:

# same operation we do to convert tuple to array--> np.asarray()

Analysing numpy array

In [61]:

# shape

x=np.random.randint(10,99,(4,5))
print(x)
print("Shape of the array:\n ",x.shape)

[[74 11 89 32 93]
[95 25 48 50 75]
[47 52 63 36 52]
[78 39 15 84 37]]
Shape of the array:
(4, 5)

In [62]:

# dimenation of the array

x.ndim

Out[62]:

In [64]:

# number of the element present in the array

x.size

Out[64]:

20

In [65]:

#checking the datatype of the element in the array

x.dtype

Out[65]:

dtype('int32')

In [100]:

myarr=np.random.randint(10,20,10)
print(myarr)

[19 15 19 16 11 18 14 19 16 18]

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 7/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [101]:

myarr.max()

Out[101]:

19

In [102]:

# index location of max value

myarr.argmax()

Out[102]:

In [103]:

myarr.min()

Out[103]:

11

In [104]:

#index location of min value

myarr.argmin()

Out[104]:

Mathematical Opearation with NumPy array

In [66]:

list1=[1,2,3,4,5]

list2=[6,7,8,9,10]

list1+list2 #==> instead of adding this PLus operator will concatenate the two list

Out[66]:

[1, 2, 3, 4, 5, 6, 7, 8, 9, 10]

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 8/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [80]:

a=np.random.randint(10,20,(3,3))
b=np.random.randint(20,30,(3,3))

print(a)
print(b)

[[16 18 11]
[16 18 10]
[15 13 10]]
[[25 29 29]
[20 20 26]
[26 29 27]]

In [81]:

# let do some basic Mathematic Operations

print(a+b) #==> same result could be obtained by np.add(a,b)

[[41 47 40]
[36 38 36]
[41 42 37]]

In [82]:

print(a-b) #==> same result could be obtained by np.substract(a,b)

[[ -9 -11 -18]
[ -4 -2 -16]
[-11 -16 -17]]

In [83]:

print(a*b) #==> same result could be obtained by np.multiply(a,b)

[[400 522 319]

[320 360 260]
[390 377 270]]

In [84]:

print(a/b) #==> same result could be obtained by np.division(a,b)

[[0.64 0.62068966 0.37931034]

[0.8 0.9 0.38461538]
[0.57692308 0.44827586 0.37037037]]

In [135]:

arr=np.arange(0,11)
arr

Out[135]:

array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10])

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 9/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [137]:

arr+arr

Out[137]:

array([ 0, 2, 4, 6, 8, 10, 12, 14, 16, 18, 20])

In [138]:

arr/arr

C:\Users\SSRVC\AppData\Local\Temp\ipykernel_18212\1862401812.py:1: Runtime
Warning: invalid value encountered in true_divide
arr/arr

Out[138]:

array([nan, 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.])

In [139]:

# since there is zero values at first place at both array which make no sence with math p
#in normal python it could be error where as in array function it shows nan values at tha

In [140]:

## there are some universal function also

In [141]:

arr

Out[141]:

array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10])

In [142]:

np.sqrt(arr)

Out[142]:

array([0. , 1. , 1.41421356, 1.73205081, 2. ,

2.23606798, 2.44948974, 2.64575131, 2.82842712, 3. ,
3.16227766])

In [143]:

np.sin(arr)

Out[143]:

array([ 0. , 0.84147098, 0.90929743, 0.14112001, -0.7568025 ,

-0.95892427, -0.2794155 , 0.6569866 , 0.98935825, 0.41211849,
-0.54402111])

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 10/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [144]:

np.log(arr)

C:\Users\SSRVC\AppData\Local\Temp\ipykernel_18212\3120950136.py:1: Runtime
Warning: divide by zero encountered in log
np.log(arr)

Out[144]:

array([ -inf, 0. , 0.69314718, 1.09861229, 1.38629436,

1.60943791, 1.79175947, 1.94591015, 2.07944154, 2.19722458,
2.30258509])

In [145]:

# since log of zero is infinity but still it giving result and infity at its place which

In [146]:

arr

Out[146]:

array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10])

In [147]:

arr.min()

Out[147]:

In [148]:

arr.max()

Out[148]:

10

In [149]:

arr.mean()

Out[149]:

5.0

In [150]:

arr.std()

Out[150]:

3.1622776601683795

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 11/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [151]:

arr.var()

Out[151]:

10.0

In [152]:

arr=np.random.randint(1,10,(5,5))
arr

Out[152]:

array([[1, 9, 4, 2, 2],
[3, 1, 5, 1, 9],
[6, 7, 4, 1, 4],
[9, 6, 3, 3, 9],
[6, 8, 9, 1, 9]])

In [153]:

# summation across the rows

arr.sum(axis=0)

Out[153]:

array([25, 31, 25, 8, 33])

In [154]:

# summation across the column

arr.sum(axis=1)

Out[154]:

array([18, 19, 22, 30, 33])

Array Manipulation

In [87]:

# transpose

array1=np.random.randint(10,20,(2,3))

print(array1)
array1.shape

[[14 12 19]
[11 12 19]]

Out[87]:

(2, 3)

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 12/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [88]:

trans=np.transpose(array1)
print(trans)
trans.shape

[[14 11]
[12 12]
[19 19]]

Out[88]:

(3, 2)

In [90]:

# reshaping array
a=np.random.randint(10,30,(2,3))
print(a)
print(a.shape)

[[15 23 18]
[25 18 13]]
(2, 3)

In [91]:

b=a.reshape(3,2)

print(b)
print(b.shape)

[[15 23]
[18 25]
[18 13]]
(3, 2)

Numpy Indexing and selection

Grabbing single element
Grabbing a slice of element
Brodcasting selection
Indexing and selection in two dimentions
Condition Selection

Grabbing single element

In [105]:

myarr=np.arange(0,11)
print(myarr)

[ 0 1 2 3 4 5 6 7 8 9 10]

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 13/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [106]:

myarr[3]

Out[106]:

In [107]:

myarr[3:5] #==> five is excluding

Out[107]:

array([3, 4])

In [108]:

myarr[:5]

Out[108]:

array([0, 1, 2, 3, 4])

In [109]:

myarr[5:]

Out[109]:

array([ 5, 6, 7, 8, 9, 10])

Brodcasting

In [110]:

myarr

Out[110]:

array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10])

In [111]:

# it means you can reassign multiple values simultaneously

myarr[0:5]=100

In [112]:

myarr

Out[112]:

array([100, 100, 100, 100, 100, 5, 6, 7, 8, 9, 10])

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 14/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [113]:

arr=np.arange(0,11)
arr

Out[113]:

array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10])

slicing

In [114]:

# slicing section of array and setting it to new variable

In [115]:

arr

Out[115]:

array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10])

In [116]:

slice_of_array=arr[0:5]
slice_of_array

Out[116]:

array([0, 1, 2, 3, 4])

In [117]:

slice_of_array[:]=99
slice_of_array

Out[117]:

array([99, 99, 99, 99, 99])

In [118]:

arr

Out[118]:

array([99, 99, 99, 99, 99, 5, 6, 7, 8, 9, 10])

In [119]:

# so here it affect to its original array in order to not happen it we can do the copy of

In [120]:

arr_copy=arr.copy()

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 15/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [121]:

arr_copy[:]=100

In [122]:

arr_copy

Out[122]:

array([100, 100, 100, 100, 100, 100, 100, 100, 100, 100, 100])

In [123]:

arr

Out[123]:

array([99, 99, 99, 99, 99, 5, 6, 7, 8, 9, 10])

slicing 2D array

In [159]:

arr = np.array([[1, 2, 3, 4, 5], [6, 7, 8, 9, 10]])

print(arr[1, 1:4])

[7 8 9]

negative slicing

In [158]:

arr = np.array([1, 2, 3, 4, 5, 6, 7])

print(arr[-3:-1])

[5 6]

indexing and selection in two dimentional array

In [124]:

arr_2d=np.random.randint(10,99,(3,3))
arr_2d

Out[124]:

array([[29, 93, 17],

[52, 78, 60],
[70, 19, 97]])

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 16/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [125]:

arr_2d.shape

Out[125]:

(3, 3)

In [126]:

# to grab first row

arr_2d[0]

Out[126]:

array([29, 93, 17])

In [127]:

# to grab third element in the first row which is 41

arr_2d[0][2]

Out[127]:

17

In [128]:

# to get some slice or subsection of matrix

arr_2d

Out[128]:

array([[29, 93, 17],

[52, 78, 60],
[70, 19, 97]])

In [129]:

arr_2d[:2,1:]

Out[129]:

array([[93, 17],
[78, 60]])

In [156]:

arr = np.array([[1,2,3,4,5], [6,7,8,9,10]])

print('2nd element on 1st row: ', arr[0, 1])

2nd element on 1st row: 2

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 17/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

Indexing in 3D array

In [157]:

import numpy as np

arr = np.array([[[1, 2, 3], [4, 5, 6]], [[7, 8, 9], [10, 11, 12]]])

print(arr[0, 1, 2])

Condition and selection

In [130]:

arr=np.arange(0,11)
arr

Out[130]:

array([ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10])

In [131]:

arr>4

Out[131]:

array([False, False, False, False, False, True, True, True, True,

True, True])

In [132]:

bool_arr=arr>4

In [133]:

arr[bool_arr] #==> filtering

Out[133]:

array([ 5, 6, 7, 8, 9, 10])

In [134]:

# same result can be obtained by following code

arr[arr>4]

Out[134]:

array([ 5, 6, 7, 8, 9, 10])

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 18/19

7/3/23, 12:41 PM Numpy Array - Complete Tutorial - Jupyter Notebook

In [163]:

print("="*100)

==========================================================================
==========================

In [ ]:

localhost:8888/notebooks/Desktop/Data Science/Python Practice/Numpy Array - Complete Tutorial.ipynb 19/19