Gain insights into utilizing NumPy for data manipulation and analytics. Learn about implementing concepts in both Python and NumPy through coding challenges and quizzes.

python.tar.gz

python3.6

Live3.6

If you're looking to grow your career in machine learning or data science in this day and age, adding a powerful library to your skill set is an important place to start. In that vein, Python has become one of the most widely used tools in the industry for serious data analytics, and NumPy is probably the most widely used data analytics library. With NumPy, you can manipulate data involving multi-dimensional arrays and matrices (think linear algebra).

Join us as we venture into the vast world of NumPy in this comprehensive course. Each lesson dive into the actual implementation of concepts in both pure Python and then NumPy, exploring how NumPy vectorization compares to traditional Python that uses a procedural and object-oriented approach.

Practice and test yourself along the way with in-browser coding challenges, quizzes, and more.

This course is intended for users who are already familiar with intermediate level Python.

From Python to Numpy

Let's consider the unit surface and a minimum radius `r` to be enforced between each point.

Knowing that the densest packing of circles in the plane is the hexagonal lattice of the bee's honeycomb, we know this density is $d = \frac{1}{6}\pi\sqrt{3}$ (in fact [I learned it](https://en.wikipedia.org/wiki/Circle_packing) while writing this course). Considering circles with radius `r`, we can pack at most $\frac{d}{\pi r^2} = \frac{\sqrt{3}}{6r^2} = \frac{1}{2r^2\sqrt{3}}$.



We know the theoretical upper limit for the number of discs we can pack onto the surface, but we'll likely not reach this upper limit because of random placements.


Furthermore, because a lot of points will be rejected after a few have been accepted, we need to set a limit on the number of successive failed trials before we stop the whole process.

$$$$



# Python Implementation

In this lesson, we will try to do the blue noise sampling using the DART method discussed in the previous lesson. We will look at both solutions, i.e., Pythonic and NumPy approach and see which one is more efficient. 

Introduction

Anatomy of an Array

Code Vectorization

Problem Vectorization

Custom Vectorization

Beyond NumPy

Conclusion

Coding Example: Blue Noise Sampling using DART method