1. Software for data analysis

Install Windows Terminal

Go to Windows store and install the Windows Terminal.

Install WSL

• Follow the Microsoft recipe.
• Open the Windows Terminal and type

wsl --install

Install Ubuntu 22.04

• Open the Windows Terminal and type

wsl --status # verify you have WSL 2 installed
wsl --list --online # check available OS to you
wsl --install --d Ubuntu-22.04 # install Ubuntu 22.04
# If it does not work, try wsl --update. You might need a proxy.

• Alternatively, you may also download it from the Windows Store directly.
• If your Windows store is blank, read this article.

Verify installation

• Verify you got Ubuntu 22.04 virtual machine.

wsl -l # verify you have Ubuntu-22.04 listed and marked as default.
wsl --set-default Ubuntu-22.04 # if not default, use this command

• Now you can enter your Ubuntu 22.04 virtual machine by
  • Open Windows Terminal and choose Ubuntu 22.04 profile
  • Or Open Windows Terminal and choose Command Prompt and type

  wsl
  

Install python

• For Mac, open iTerm2 and use brew

brew install python

• You can also use default python of Mac, but I recommend you use the new python from brew.

• For windows, open Windows Terminal and open Ubuntu 22.04 (WSL)

sudo apt update
sudo apt upgrade
sudo apt install python3

Setup python virtual-env

• python program always uses some libraries, and sometimes different library depend on inconsistent versions of the same library and this would breaks python.
• This can be largely avoided by use of virtual-env.
• A virtual-env is an isolated folder supposed to contain all library files you need for some project.
• I recommend you always use virtual-env.
• Alternatively you can use miniconda, but I had very bad experience with it. Please remove it if you have installed it in your Ubuntu or Mac.
• You can keep your miniconda or Anaconda on your Windows. We don’t work in the Windows directly.
• Now create the venv in some folders, say ~/software/python-venv/physics

python -m venv ~/software/python-venv/physics

• Then activate it

source ~/software/python-venv/physics/bin/activate

• To avoid typing the above every time you open a new shell,
  • For Ubuntu, put it in the ~/.bashrc.
  • For Mac, put it in the ~/.zshrc.

Install jupyter etc.

pip install jupyter matplotlib pandas numpy scipy mplhep uproot pyarrow iminuit

Install VSCode

• Download the Visual Studio Code from its official website.
  • It’s better than Jetbrains CLion regarding support for remote development.
  • But CLion has better support for perf. If you know how to perf with VSCode, email me and let me know.
  • Now we can stick with VSCode.
• Follow the stackoverflow post so you can launch VSCode from your shell. It’s convenient.

Start jupyter in VSCode

• Use your shell to create a folder

mkdir -p ~/work/physics/tutorial

• Open VSCode and select from the menu File -> Open Folder ..., then choose the folder you just created ~/work/physics/tutorial.

  • Alternatively you can

  cd ~/work/physics/tutorial
  code .
  

• Create a new file named test.ipynb

• Now click on the right top and select the python you use for jupyter. It should be the one in the venv earlier.

• Now type in one cell

print("hi")

• Press CTRL+Enter, now you should run the cell and see hi.

Run iminuit tutorial

• iminuit is the python version of the famous Minuit software library.
• Now you just copy and paste code on iminuit tutorial and test that you setup all softwares.
• Put this in one cell

# basic setup of the notebook
from matplotlib import pyplot as plt
import numpy as np

# everything in iminuit is done through the Minuit object, so we import it
from iminuit import Minuit

# we also need a cost function to fit and import the LeastSquares function
from iminuit.cost import LeastSquares

# display iminuit version
import iminuit
print("iminuit version:", iminuit.__version__)
# our line model, unicode parameter names are supported :)
def line(x, α, β):
    return α + x * β


# generate random toy data with random offsets in y
rng = np.random.default_rng(1)
data_x = np.linspace(0, 1, 10)
data_yerr = 0.1  # could also be an array
data_y = rng.normal(line(data_x, 1, 2), data_yerr)
least_squares = LeastSquares(data_x, data_y, data_yerr, line)

m = Minuit(least_squares, α=0, β=0)  # starting values for α and β

m.migrad()  # finds minimum of least_squares function
m.hesse()   # accurately computes uncertainties

• Press CTRL + Enter and you should see the plots