MCandPython

Jupyter sessions for lecture on Python and Monte Carlo techniques.

1. Introduction to Python for numerical calculations

   • The language: basics and syntax
   • Numpy module
   • Matplotlib module

2. Monte Carlo techniques in statistics and error handling.

   • Estimator
   • probability density functions
   • Monte Carlo methods
   • Error propagation

How to install Python and that is necessary for the lectures

• What is mandatory:
  • Python > v3.5
  • jupyter module
  • numpy module
  • matplotlib module
  • scipy module

Windows:

• If you are familiar with prompts or Python is already installed by administrators but you can not have admin access go to (ii.):

  1. Go to https://www.python.org/ and download the windows version (or go directly here: https://www.python.org/downloads/windows/ ). Take executable installer of the version > 3.5 (currently at the creation of this readme it is 3.6.1)

  2. Open a powershell prompt (search for powershell executable)

  3. Then create a "virtual environment" that will have isolated environment in which you can install modules without affecting the main installation:

     python -m venv myenv
     

  4. Set the environment variables:

     myenv\Scripts\activate
     

  5. Now install jupyter module :

     pip3 install jupyter
     

  6. Now install the other modules :

     pip3 install numpy
     pip3 install matplotlib
     pip3 install scipy
     pip3 install pandas
     

  7. You can start a jupyter server with: jupyter notebook

• If you have admin access and want to be quick :
  1. Go to Anaconda website: https://www.continuum.io/downloads and download the packaged python distribution that will includes jupyter, numpy, etc in one setup.

Linux :

1. Install through your distribution the version python that you want (2.7 or >3.5)

2. Use pip install or pip install --user to install the following modules :

   pip3 install --user jupyter
   pip3 install --user numpy
   pip3 install --user scipy
   pip3 install --user pandas
   pip3 install --user matplotlib
   

3. You can start a jupyter server with

   jupyter notebook
   

Acknowledgments

Those seminar sessions were written by Christophe Rappold

License

Those sessions are published under the terms of the CC BY-SA 4.0 license