Overview

This cheatsheet covers the minimal key aspects on how to use "the Numpy Stack". This stack is foundational to any machine learning and data science work done in Python.

The cheat sheet covers:

1. Python: the absolute bare minimum you need to know to do something useful.
2. Numpy: provides the basic numpy array abstraction to work with multi-dimensional arrays in Python. Everything else is built upon this.
3. Matplotlib: A library for data visualization.
4. Pandas: a library for reading, writing and manipulating data.
5. SciPy: statistical computations (among many other things).

The work is based on a mixture of different resources. Notably:

• https://www.udemy.com/course/deep-learning-prerequisites-the-numpy-stack-in-python/learn/lecture/19643200#overview
• http://ai.berkeley.edu/tutorial.html

Table of Content

Part 1- Python

• Running Modes
• Getting help
• Strings
• Data Structures
  • Lists
  • Tuples
  • Sets
  • Dictionaries: Key-value map
  • Iteration
    • List Comprehensions
  • Sorting iterables
• If statements
• Functions
• Classes
  • Using Classes
• Unit Testing
• Debugging

Part 2 - Numpy

• NP array creation
• Adding elements to an array
• Array access
• Iteration
• Scalar operations
• Element-wise application of standard mathematical functions
  • Element-wise application of arbitrary functions and lambdas
• Vector to Vector and Matrix to Matrix element-wise operations
• Vector operations
  • Dot Product of Vectors
  • Magnitude (aka norm) and cosine similarity of vectors
• Matrix operations
  • Solving linear systems
• Generating Data
  • Generating Random Numbers with the np.random module
• Descriptive statistics of arrays
• Numpy array axes and operations along axes
  • Operations along axes
  • Sum, mean, standard deviation, max, argmax, and others
  • Concatenate
  • Choosing a random number
• Sorting
• Searching

Part 3 - Matplotlib

• Line charts
• Scatter plots
• Histograms
  • Simple histogram
  • Overlapping Histograms
• Plotting images
  • Converting images to black and white
• Heatmaps
• Combined Plots

Part 4 - Pandas

• Reading Data
• Dataframe indexes
• Understanding the structure of the data
  • Replacing column names in a dataframe
• Select Rows and Columns in a dataframe
• Converting Dataframes into Numpy Arrays
• Writing to a file
• Applying / Mapping a function to each row in a dataframe
• Get descriptive statistics of the columns
• Plotting with Pandas
  • Histogram
  • Line chart
  • Box Plot
  • Scatter matrix

Part 5 - SciPy

• Probability Density Functions (PDF) and Cumulative Density Functions (CDF)
• Convolution