Introduction to Jupyter Notebook, Numpy, & Matplotlib

• Numpy+Scipy+Matplotlib=Matlab
• AI in Geosciences
• Computational Interpretation Group
• University of Science and Technology of China
• Xinming Wu
• 10/09/2020

Outline

Jupyter notebook

Numpy

Matplotlib

Jupyter Notebook

What is Jupyter Notebook

• open-source web application: https://jupyter.org/

• create and share documents that contain live code, equations, visualizations and narrative text.

• data cleaning and transformation, numerical simulation, statistical modeling, data visualization, machine learning, and much more.

• From when you start a Jupyter notebook, variable assignments, function declarations, etc are stored and updated as you go.

Useful features

• Hit shift+enter to execute the cell

• Use ! to excute shell commands, try !ls, !pwd, !pip install ...

• Use ? (e.g., ?numpy, ?numpy.mean()) for quick reference on syntax

• Change themes
  • pip install jupyterthemes to install themes
  • jt -l to check the available themes
  • jt -t <theme> to set the theme
  • jt -r back to the default theme

• Notebook Extensions
  • install: type pip install jupyter_contrib_nbextensions and then jupyter contrib nbextension install
  • Table of Contents
  • Hinterland
  • Split cells

• Using LaTeX for forumlas
  • $\frac{\partial^2u}{\partial t^2}=c^2\frac{\partial^2 u}{\partial x^2}$

• Multicursor support

Sharing your notebooks

• Before you share
  • Click “Cell > All Output > Clear”
  • Click “Kernel > Restart & Run All”
  • Wait for your code cells to finish executing and check ran as expected

• Export your notebooks
  • Convert notebooks to html files using the File > Download as > HTML Menu option.
  • Create a presentation like this one by using RISE
  • jupyter nbconvert numpy.ipynb --to slides --SlidesExporter.reveal_scroll=True
  • Use the File > Download as > PDF menu to save your notebook as a PDF. (http://blog.juliusschulz.de/blog/ultimate-ipython-notebook#converting-notebook-to-latex-pdf)

Numpy

Array programming with NumPy

Nature volume 585, pages 357–362 (2020)

Charles R. Harris, K. Jarrod Millman, and Stéfan J. van der Walt et al.

What is Numpy (Numerical Python)

• a fundamental package for scientific computing in Python
• a Python library includes a multidimensional array object (ndarray), various derived objects, and fast operations on arrays
• one of the top five Python packages, used in every field of science and engineering.
• install numpy: pip install numpy
• import numpy: import numpy as np

Why is Numpy

• fast
• the core of the scientific Python and PyData ecosystems
• the universal standard for working with numerical data in Python
• works well with Pandas, SciPy, Matplotlib, scikit-learn, scikit-image and most other data science and scientific Python packages
• all kinds of matrix operations, lots of built-in functions

ndarray: the core of NumPy

• a homogeneous n-dimensional array object (unlike Python sequences)
  • The elements in a NumPy array are of the same data type and the same size in memory.
• NumPy arrays have a fixed size at creation, unlike Python lists (which can grow dynamically)
• a NumPy array has continuous memory

A 1D array is a vector

its shape is just the number of components

import numpy as np

data = np.array([1,2,3])
print(data)

[1 2 3]

A 2D array is a matrix; its shape is (number of rows, number of columns)

data = np.array([[1,2],
               [3,4],
               [5,6]])
print(data)

[[1 2]
 [3 4]
 [5 6]]

Structure of a 3D (3, 4, 2) array

Memory and strides: what make numpy fast

• a NumPy array is stored in a contiguous block of memory
• two key concepts relating to memory: dimensions and strides
• strides are integer numbers of bytes to step in each dimension

An example of 2D array

data = np.array([[1,2,3],[4,5,6]], dtype='int16')
data.strides

(6, 2)

shape of the 2D array

continuous memory and strides

The byte address of an element data[i1,i2]:

byte_address = data.strides[0] * i1 + data.strides[1] * i2

2D array shape and memory