Computational Economics with Python

Graduate Mini Course at Columbia University

• Instructor: John Stachurski
• Dates: 26--28th March 2018
• Times and location: see here

Summary

This mini course will provide a fast paced introduction to Python for computational economic modeling, from basic scripting to high performance computing. The course is aimed at graduate students with proficiency in at least one scientific computing platform (e.g, MATLAB, Fortran, STATA, R, C or Julia).

No Python knowledge is assumed.

Please be sure to bring your laptop

Instructions

Get Python + scientific libraries

• Install Anaconda Python

Update Numba (still necessary as of 25th March 2018)

• At terminal (Mac / Linux) or Anaconda Prompt (Windows), type conda install numba=0.37

Get files from this repo

• Use git clone if you know git or download the zip file

Schedule

Day 1

• Python vs MATLAB vs Julia vs Fortran vs others
• The Python language: syntax and semantics
• Object oriented vs procedural programming
• Jupyter notebooks

Day 2

• The major scientific libraries ( SciPy / NumPy / Matplotlib / etc.)
• Numba and other JIT compilers
• Parallelization
• Distributed and cloud computing

Day 3

• Applications (asset pricing, optimal savings, optimal stopping)

Links:

• Anaconda
• AWS
• Accessing AWS via SSH

Notes on AWS

To get an instance running

1. Login to Amazon AWS Console
2. Navigate to EC2 Service
3. Choose your region for setting up an instance
4. Create security key-pair for the region if you don't have one
5. Launch & Configure an instance and choose Ubuntu 64-bit
6. enable access through Port 8000 (in addition to Port 22 for ssh)
7. Choose security key you've set up

Connecting and set up

Use ssh -i /path/to/pem-key ubuntu@hostname

Here hostname is your Public DNS, as shown in the instance information from AWS console

Now run sudo apt-get update so you can install things you might need using apt-get

Configure instance to run Jupyter

1. ssh into the running instance using IP from AWS Console
2. Install Anaconda using wget and the latest download link for python36
3. Run: jupyter notebook --generate-config
4. For Automatic Password Setup run: jupyter notebook password
5. Edit .jupyter/jupyter_notebook_config.py and set the following

# Set ip to '*' to bind on all interfaces (ips) for the public server
c.NotebookApp.ip = '*'
c.NotebookApp.open_browser = False
c.NotebookApp.port = 8000