Numpy - Part 1

💡Why Numpy ?

 1️⃣ Numerical Operations are faster than normal python lists

 2️⃣ Also known as Array Oriented Computing

 3️⃣ Unlike lists, Numpy Array elements should be of same data type

 4️⃣ package for creating Multi-Dimension Arrays

some info on Arrays

18 is a Scalar (there is no dimension, so no Shape)

1️⃣ D Array = Vector (collection of scalars) [1, 2, 3, 4, 5]

It has 1 axes i.e., axis = 0

Shape is (N, ) - N is no. of scalars

here, Shape is (5, )

2️⃣ D Array = Matrix (collection of Vectors) [[1, 2, 3, 4, 5], [6, 7, 8, 9, 10]]

It has 2 axes i.e., axis=0, axis=1

Shape is (N, M) - N is no. of vectors, M is no of scalars in each vector

here, Shape is (2, 5)

3️⃣ D Array (collection of Matrix's)

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

anything more than 2D array can be called Tensor

Shape is (N, M, P) - N number of Matrices each of shape (M, P)

here, Shape is (2, 3, 5)

💡 How to Create an Array ?

3 ways

import numpy as np

➖ np.array(your_list) your list or nested list to create 1D/higher dimension
arrays

➖ np.arange(5) - creates 1D Array read it as a range (range() in python)

➖ np.random.rand(5)

💡 Understand our Array

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

→ a.dtype # data type is dtype('int64') → a.dim # dimension is 1 → a.shape # shape is (4, ) → a.size # total number of elements

If you like the content, you could follow on twitter @nameismahipal

Numpy - Part 1

Comments

Machine Learning

More from this blog

Git Commands

CSS - Inheritance and Others

Web Browser Engine Overview : For Beginners

git - part 2: for Beginners

Command Palette

Comments

Machine Learning

More from this blog