Using Numpy in Python

Hello all,

This is the third article in the series Python for Data Science. If you are new to this series, we would recommend you to read our previous articles

  1. Python for Data Science Series - Part 1
  2. Python for Data Science Series - Part 2


To install numpy in your python environment, please use the below command

pip install numpy


Please refer the video below for detailed explanation on Numpy


After you have installed numpy, please refer the following notebook to understand on how to use Numpy functionalities.




In [1]:
import numpy as np

Defining a Numpy Array

In [2]:
a = np.array([10,20,30])
In [3]:
a
Out[3]:
array([10, 20, 30])
In [4]:
type(a)
Out[4]:
numpy.ndarray
In [ ]:
dir(a)
In [6]:
a.shape
Out[6]:
(3,)
In [7]:
a = np.array([[1,2,4],[5,6,7]])
In [8]:
a.shape
Out[8]:
(2, 3)
In [9]:
a
Out[9]:
array([[1, 2, 4],
       [5, 6, 7]])
In [10]:
a = np.array([[[1,2,4],[5,6,7]],[[1,2,10],[8,4,2]]])
In [11]:
a.shape
Out[11]:
(2, 2, 3)
In [12]:
a
Out[12]:
array([[[ 1,  2,  4],
        [ 5,  6,  7]],

       [[ 1,  2, 10],
        [ 8,  4,  2]]])
In [13]:
type(a)
Out[13]:
numpy.ndarray
In [14]:
a.dtype
Out[14]:
dtype('int32')
In [15]:
b = a.T
b
Out[15]:
array([[[ 1,  1],
        [ 5,  8]],

       [[ 2,  2],
        [ 6,  4]],

       [[ 4, 10],
        [ 7,  2]]])
In [16]:
b.shape
Out[16]:
(3, 2, 2)
In [17]:
a = np.array([[1,2],[3,4]])
In [18]:
a
Out[18]:
array([[1, 2],
       [3, 4]])
In [19]:
a.shape
Out[19]:
(2, 2)
In [20]:
b = a.T
b
Out[20]:
array([[1, 3],
       [2, 4]])
In [21]:
a = np.array([1,103.34,34])
In [22]:
a.dtype
Out[22]:
dtype('float64')

MIN MAX ARGMIN ARGMAX

In [23]:
max(a)
Out[23]:
103.34
In [24]:
b
Out[24]:
array([[1, 3],
       [2, 4]])
In [25]:
np.max(b)
Out[25]:
4
In [26]:
np.argmax(b)
Out[26]:
3
In [27]:
a = np.array([1,2,3,4])
In [28]:
np.argmax(a)
Out[28]:
3

Generating Arrays

In [31]:
np.ones(shape=(2,3))
Out[31]:
array([[1., 1., 1.],
       [1., 1., 1.]])
In [32]:
np.zeros(shape=(3,3))
Out[32]:
array([[0., 0., 0.],
       [0., 0., 0.],
       [0., 0., 0.]])
In [40]:
np.eye(3)
Out[40]:
array([[1., 0., 0.],
       [0., 1., 0.],
       [0., 0., 1.]])
In [44]:
np.random.sample(4)
Out[44]:
array([0.07908825, 1.83392063, 5.07344989, 1.90522311])
In [1]:
a = list(range(0,100))
In [58]:
np.random.choice
Out[58]:
<function RandomState.choice>
In [61]:
np.random.sample((5,5))
Out[61]:
array([[0.67240121, 0.03435943, 0.0911954 , 0.22860258, 0.72933351],
       [0.20351709, 0.76852415, 0.47024114, 0.97237468, 0.18860797],
       [0.94120166, 0.12416461, 0.04580034, 0.15012586, 0.67994007],
       [0.82487477, 0.08379525, 0.10066431, 0.96580346, 0.62604015],
       [0.51248987, 0.81711833, 0.67205623, 0.42366449, 0.59213291]])

Matirx Operations

In [62]:
a = np.array([[1,2],[3,4]])
a
Out[62]:
array([[1, 2],
       [3, 4]])
In [63]:
b = np.array([[5,6],[7,8]])
b
Out[63]:
array([[5, 6],
       [7, 8]])
In [64]:
a + b
Out[64]:
array([[ 6,  8],
       [10, 12]])
In [65]:
a * b
Out[65]:
array([[ 5, 12],
       [21, 32]])
In [66]:
np.matmul(a,b)
Out[66]:
array([[19, 22],
       [43, 50]])
In [67]:
np.add(a,b)
Out[67]:
array([[ 6,  8],
       [10, 12]])
In [71]:
np.subtract(a,b)
Out[71]:
array([[-4, -4],
       [-4, -4]])
In [72]:
a
Out[72]:
array([[1, 2],
       [3, 4]])
In [73]:
a.T
Out[73]:
array([[1, 3],
       [2, 4]])
In [83]:
a = np.array([[[1,2,3],[4,5,6],[7,8,9]],[[11,12,3],[14,15,16],[17,18,19]]])
In [84]:
a
Out[84]:
array([[[ 1,  2,  3],
        [ 4,  5,  6],
        [ 7,  8,  9]],

       [[11, 12,  3],
        [14, 15, 16],
        [17, 18, 19]]])
In [85]:
a.shape
Out[85]:
(2, 3, 3)

Slicing Arrays

In [89]:
a[1][0][0]
Out[89]:
11
In [103]:
a[1,0,:]
Out[103]:
array([11, 12,  3])

Transpose Arrays

In [107]:
b = np.transpose(a)
b
Out[107]:
array([[[ 1, 11],
        [ 4, 14],
        [ 7, 17]],

       [[ 2, 12],
        [ 5, 15],
        [ 8, 18]],

       [[ 3,  3],
        [ 6, 16],
        [ 9, 19]]])
In [108]:
b.shape
Out[108]:
(3, 3, 2)
In [115]:
b[2,:,1]
Out[115]:
array([ 3, 16, 19])
In [118]:
a = np.array([[[1,2,3,100],[4,5,6,200],[7,8,9,300]],[[11,12,3,400],[14,15,16,500],[17,18,19,600]]])
a
Out[118]:
array([[[  1,   2,   3, 100],
        [  4,   5,   6, 200],
        [  7,   8,   9, 300]],

       [[ 11,  12,   3, 400],
        [ 14,  15,  16, 500],
        [ 17,  18,  19, 600]]])
In [119]:
a.shape
Out[119]:
(2, 3, 4)
In [122]:
b = np.transpose(a)
b
Out[122]:
array([[[  1,  11],
        [  4,  14],
        [  7,  17]],

       [[  2,  12],
        [  5,  15],
        [  8,  18]],

       [[  3,   3],
        [  6,  16],
        [  9,  19]],

       [[100, 400],
        [200, 500],
        [300, 600]]])
In [123]:
b.shape
Out[123]:
(4, 3, 2)
In [126]:
b = np.transpose(a,axes=(0,2,1))
b
Out[126]:
array([[[  1,   4,   7],
        [  2,   5,   8],
        [  3,   6,   9],
        [100, 200, 300]],

       [[ 11,  14,  17],
        [ 12,  15,  18],
        [  3,  16,  19],
        [400, 500, 600]]])
In [127]:
b.shape
Out[127]:
(2, 4, 3)

Concatenation of Arrays

In [128]:
a = [1,2,3,4]
In [131]:
np.asarray(a)
Out[131]:
array([1, 2, 3, 4])
In [136]:
a = np.array([1,2,3,4])
b = np.array([5,6,7,8])
In [137]:
np.append(a,b)
Out[137]:
array([1, 2, 3, 4, 5, 6, 7, 8])
In [140]:
c = np.hstack((a,b))
c
Out[140]:
array([1, 2, 3, 4, 5, 6, 7, 8])
In [141]:
d = np.vstack((a,b))
d
Out[141]:
array([[1, 2, 3, 4],
       [5, 6, 7, 8]])
In [142]:
c.shape
Out[142]:
(8,)
In [143]:
d.shape
Out[143]:
(2, 4)
In [148]:
a = np.random.random((2,3))
print(a.shape)
a

(2, 3)
Out[148]:
array([[0.89779757, 0.70373201, 0.55522516],
       [0.76259986, 0.15352072, 0.28630461]])
In [149]:
b = np.random.random((2,3))
print(b.shape)
b

(2, 3)
Out[149]:
array([[0.94099365, 0.11215319, 0.0237094 ],
       [0.82909191, 0.00599765, 0.683048  ]])
In [150]:
c = np.hstack((a,b))
print(c.shape)
c

(2, 6)
Out[150]:
array([[0.89779757, 0.70373201, 0.55522516, 0.94099365, 0.11215319,
        0.0237094 ],
       [0.76259986, 0.15352072, 0.28630461, 0.82909191, 0.00599765,
        0.683048  ]])
In [151]:
d = np.vstack((a,b))
print(d.shape)
d

(4, 3)
Out[151]:
array([[0.89779757, 0.70373201, 0.55522516],
       [0.76259986, 0.15352072, 0.28630461],
       [0.94099365, 0.11215319, 0.0237094 ],
       [0.82909191, 0.00599765, 0.683048  ]])
In [ ]:
In [ ]:
In [152]:
a = np.random.random((2,3,4))
print(a.shape)
a

(2, 3, 4)
Out[152]:
array([[[0.05348472, 0.38735588, 0.07242871, 0.37705306],
        [0.10991526, 0.68119907, 0.700878  , 0.81107797],
        [0.0624925 , 0.99382657, 0.82772526, 0.61071547]],

       [[0.9282276 , 0.42039079, 0.94785463, 0.69754789],
        [0.41982185, 0.0625692 , 0.4286561 , 0.69704102],
        [0.92536563, 0.51611501, 0.23019467, 0.7254887 ]]])
In [153]:
b = np.random.random((2,3,4))
print(b.shape)
b

(2, 3, 4)
Out[153]:
array([[[0.51753157, 0.12818528, 0.37229487, 0.01687935],
        [0.00573693, 0.80945614, 0.5601282 , 0.88981313],
        [0.49319538, 0.92671174, 0.96977514, 0.41044506]],

       [[0.553852  , 0.08528831, 0.63869774, 0.13570784],
        [0.40396164, 0.24848851, 0.67898245, 0.86057556],
        [0.28326273, 0.59721677, 0.45035408, 0.05016073]]])
In [154]:
c = np.hstack((a,b))
print(c.shape)
c

(2, 6, 4)
Out[154]:
array([[[0.05348472, 0.38735588, 0.07242871, 0.37705306],
        [0.10991526, 0.68119907, 0.700878  , 0.81107797],
        [0.0624925 , 0.99382657, 0.82772526, 0.61071547],
        [0.51753157, 0.12818528, 0.37229487, 0.01687935],
        [0.00573693, 0.80945614, 0.5601282 , 0.88981313],
        [0.49319538, 0.92671174, 0.96977514, 0.41044506]],

       [[0.9282276 , 0.42039079, 0.94785463, 0.69754789],
        [0.41982185, 0.0625692 , 0.4286561 , 0.69704102],
        [0.92536563, 0.51611501, 0.23019467, 0.7254887 ],
        [0.553852  , 0.08528831, 0.63869774, 0.13570784],
        [0.40396164, 0.24848851, 0.67898245, 0.86057556],
        [0.28326273, 0.59721677, 0.45035408, 0.05016073]]])
In [155]:
d = np.vstack((a,b))
print(d.shape)
d

(4, 3, 4)
Out[155]:
array([[[0.05348472, 0.38735588, 0.07242871, 0.37705306],
        [0.10991526, 0.68119907, 0.700878  , 0.81107797],
        [0.0624925 , 0.99382657, 0.82772526, 0.61071547]],

       [[0.9282276 , 0.42039079, 0.94785463, 0.69754789],
        [0.41982185, 0.0625692 , 0.4286561 , 0.69704102],
        [0.92536563, 0.51611501, 0.23019467, 0.7254887 ]],

       [[0.51753157, 0.12818528, 0.37229487, 0.01687935],
        [0.00573693, 0.80945614, 0.5601282 , 0.88981313],
        [0.49319538, 0.92671174, 0.96977514, 0.41044506]],

       [[0.553852  , 0.08528831, 0.63869774, 0.13570784],
        [0.40396164, 0.24848851, 0.67898245, 0.86057556],
        [0.28326273, 0.59721677, 0.45035408, 0.05016073]]])
In [162]:
e = np.random.random((1,3,4))
print(e.shape)
e

(1, 3, 4)
Out[162]:
array([[[0.4394013 , 0.72769784, 0.92034209, 0.92738419],
        [0.94154974, 0.7619238 , 0.29673614, 0.67078156],
        [0.25698888, 0.44332544, 0.99029697, 0.01074971]]])
In [163]:
f = np.vstack((d,e))
print(f.shape)
f

(5, 3, 4)
Out[163]:
array([[[0.05348472, 0.38735588, 0.07242871, 0.37705306],
        [0.10991526, 0.68119907, 0.700878  , 0.81107797],
        [0.0624925 , 0.99382657, 0.82772526, 0.61071547]],

       [[0.9282276 , 0.42039079, 0.94785463, 0.69754789],
        [0.41982185, 0.0625692 , 0.4286561 , 0.69704102],
        [0.92536563, 0.51611501, 0.23019467, 0.7254887 ]],

       [[0.51753157, 0.12818528, 0.37229487, 0.01687935],
        [0.00573693, 0.80945614, 0.5601282 , 0.88981313],
        [0.49319538, 0.92671174, 0.96977514, 0.41044506]],

       [[0.553852  , 0.08528831, 0.63869774, 0.13570784],
        [0.40396164, 0.24848851, 0.67898245, 0.86057556],
        [0.28326273, 0.59721677, 0.45035408, 0.05016073]],

       [[0.4394013 , 0.72769784, 0.92034209, 0.92738419],
        [0.94154974, 0.7619238 , 0.29673614, 0.67078156],
        [0.25698888, 0.44332544, 0.99029697, 0.01074971]]])
In [ ]:
Location: India

Comments

Post a Comment

Popular posts from this blog

Project: Implementation of Label Editor

Image
 This is second article in Python Projects Series. This article includes the explanation and instructions of demo of this project. We also touch up on the topics like design and implementation of this project. This Python code works as a label editor through which you can draw bounding boxes and label them. CSV file shall be generated from this tool which contains information about your labelling. In case if you are new to this series, Please visit our previous article in this series. 1.  Project : Implementation of Block Jump Game 2.  Project: Implementation of Car Lane Change Game   In case if you are new to Python, Please visit the following articles in series  Python for Data Science . Python for Data Science Series - Part 1 Python for Data Science Series - Part 2 Using Numpy in Python Using Pandas in Python Data Visualization using Matplotlib Python Parsing XML file in python Parsing Json file in python Interactive Data Visualization using Plotly in Python ...
Read more

Using Pandas in Python

Image
Hello all, This is the fourth article in the series  Python for Data Science . If you are new to this series, we would recommend you to read our previous articles Python for Data Science Series - Part 1 Python for Data Science Series - Part 2 Using Numpy in Python To install pandas in your python environment, please use the below command pip install pandas Please refer the videos below for detailed explanation on Pandas After you have installed Pandas, please refer the following notebook to understand on how to use Pandas functionalities. In [1]: import pandas as pd import os Creating a DataFrame ¶ In [2]: data1 = [ 1 , 2 , 3 ] data2 = [ 4 , 5 , 6 ] data3 = [ 7 , 8 , 9 ] pd . DataFrame ([ data1 , data2 , data3 ], columns = [ 'a' , 'b' , 'c' ], index = [ 'row1' , 'row2' , 'row3' ]) Out[2]: a b c ...
Read more