Overview

== “Numeric Python”
– Open Source Python Library for array-based calculations
– First realeased in 1995 as Numeric (first implementation of a Python matrix package); 2006 as NumPy
– allows easy handling of vectors, matrices or generally large multidimensional arrays
– NumPys operators and functions are optimized for multidimensional array operations and evaluate particularly efficiently
– written in C
– compatible to various Python libraries (Matplotlib, Pandas, SciPy)
– SciPy extends the power of NumPy with other useful features, such as: minimization, regression, Fourier transform

python vs matlab
→ Python is an alternative to Matlab

Numpy Applications

uses numpy

Python and Science

– The programming language Python is used very intensively in the application area of scientific research
– NumPy was designed for scientific calculations

numpy scientific domains

The ndarray data structure

– Core functionality of NumPy is based on the data structure “ndarray
– Components: a pointer to a contiguous storage area together with metadata describing the data stored in it
– All elements of an array must be of the same data type

numpy ndarrays

Technical terms


shape == Defines the dimensions in each index value (“axis”) of the array and the number of axes
strides == describe for each axis, how many bytes you have to jump in linear memory, if an index belonging to this axis is increased by 1
reshaping == Altering the shape of a provided array
slicing == Setting up smaller subarrays within a given larger array
splitting + joining == Splitting one array into many and combining multiple arrays into one single array
indexing == Setting the value of individual array elements

Multiplication

numpy multiplication

The product and further information can be found here:

https://numpy.org/