Seminar »Data Processing with NumPy« from May 22 to May 23, 2017
| When | May 22, 2017 09:00 AM
to
May 23, 2017 05:00 PM |
|---|---|
| Where | Veit Schiele Communications GmbH, Mansteinstr. 7, D-10783 Berlin |
| Contact Name | Veit Schiele |
| Contact Phone | +49 30 8185667-1 |
| Add event to calendar |  vCal iCal |
Prerequisites
Basic knowledge of Python
Target Audience
Analysts, researchers and engineers who would like to carry out numerical calculations on their data.
Course Description
NumPy is the most frequently used Python library for processing numerical data. It combines a programmer-friendly Python interface with the speed of an implementation in pure C. NumPy allows to implement calculations with large data series and matrices with few lines of code. Therefore, NumPy is a perfect fit to optimize the runtime of Python programs.
In this course you will get a hands-on introduction of NumPys essentials, featuring many practical examples. To build upon the basic functionality, the SciPy package featuring a plethora of mathematical tools that make the best use of NumPy will be covered as well.
Duration
2 days
Course Outline
| Day 1 | Day 2 |
|---|---|
| Introduction to NumPy | Broadcasting |
| Functions / ufuncs | Optimization with NumPy |
| Indexing | The Scipy Library |
| Typical Applications | Related Python Libraries |
Introduction to NumPy
- overview of the functionality in NumPy
- arrays
- dtypes
- reshape
- creating arrays
- loading/saving data
Indexing
- indexing arrays
- views
- fancy indexing
- sorting
- set operations
Functions
- built-in functions
- ufuncs
- matrix operations
- rotating coordinates
Optimization
- eliminating Python loops with NumPy
- sparse matrices
- identifying bottlenecks with cProfile
Typical Applications
- recommender systems
- Eigenvectors
- the PageRank algorithm
- implementing a neural network in NumPy
Broadcasting
- broadcasting
- stacking
- raveling
The SciPy Library
- finding zeroes
- fitting polynomial functions
- Fourier transformation
- data visualization