Skip to content

Python course for finance

Notifications You must be signed in to change notification settings

Dauphine203/py_dauphine

Repository files navigation

Course outline

  1. Basics of Python: data types, loops, functions, list comprehensions and a glimpse of objet oriented programming. (3h)
  2. Advanced Python: functional programming, generators... and an introduction to the numpy library. (3h)
  3. Data Analysis with pandas: I/O operations. Working with dates/time. Financial panel data: the pandas library. (3h)
  4. High-frequency Data and Visualisation: Data visualization the matplotlib library. (3h)
  5. Regression. Optimization: Interpolation and curve fitting. Symbolic mathematics in Python. Principal component analysis. (3h)
  6. Stochastic Processes in Python: Generating random numbers. Monte Carlo simulations. Simulating stock price paths (Brownian motion with jumps). Value-at-Risk and Expected Shortfall. (3h)
  7. Option Pricing: Option pricing with binomial trees and Monte Carlo simulation. Least-Squares Monte Carlo for pricing American options. (3h)
  8. Portfolio Theory. Efficient frontier. PCA Analysis. Test for normality. (3h)

2019-2020

Project (group of 2 or 3 people)

  • Option pricing - monte carlo method : Black Scholes diffusion is not enough + choose a convenient option for that
  • Parallelism (concurrent.futures, Joblib, Dask...)
  • Dash as a service to display web pages: Form to fill the parameters, display the pricing and draw some trajectories Expected by 6th December
  • Readme file is mandatory: how to install, how to use the app

Exam

to be planned

2020-2021

Evaluation (20 points)

  • participation (2 points)
  • first assignment (3 people) (4 points) (from 3rd session, 20 minutes presentation 10 minutes Q&A)
  • second assignment (3 people I will provide the list soon)(6 points)
  • Exam (8 points)

second assignment

  • What to do: develop a service based on Bokeh (user interface), with some calculations (options pricing, risk, whatever you want in finance) based on Dask or Cupy or Cudf Pytest are mandatory Will onboard Flake8

An article on https://medium.com/ is a bonus (e.g.: https://towardsdatascience.com/speed-cubing-for-machine-learning-a5c6775fff0b) outcome: simple package to install or a Google colab Notebook and a readme file Expected by beginning of January 2021

Releases

No releases published

Packages

No packages published