This semester we’ve written a bunch of python code. We’ve primarily written and run our programs in IDLE, but we’ve also tried running our programs from the command line. This makes a lot of sense when the point of the program is primarily to create a completed program that can then be run repeatedly.

Today I want to demonstrate another way to look at writing Python programs that makes sense for situations where:

• A core purpose of the program is to have other people look at the code in order to communicate (e.g. a scientific paper).
• The code will produce one output and the only reason to re-run it is to confirm that output.
• You’re exploring in a way that interactive programming is useful.

New tool: Jupyter Notebook

To get this set up easily, you might want to try Anaconda or Miniconda.

If you use Anaconda, then Jupyter is installed by default but you need to dodge some trick payware nonsense when you try to run it. Any sort of “cloud” thing is probably a scam.

If you use Miniconda, you can install Jupyter with

conda install jupyter

Once you have it installed, open up Jupyter Labs. This will pop up a browser window - note that this isn’t an external web site, it’s running on your machine (with a local web server, much like what we played with earlier this semester).

Let’s create a new notebook.

• Cell -> Markdown: # Hello Notebook
• Cell -> Code
  • x = 10 + 20
• Cell -> Code
  • x
• Cell -> Code

from PIL import Image
image = Image.open("duck1.jpg")
image

Exporting as PDF / HTML:

• File -> Export as: PDF
• This requires some packages to be installed.
  • conda install pandoc
  • LaTeX
    • Linux: apt-get install texlive texlive-xetex
    • Windows: choco install texlive
    • Mac: brew install texlive
• File -> Export as: HTML

And another cell:

image.rotate(180)

More Examples #

From: https://github.com/StrikingLoo/pandas_workshop

import pandas as pd
import random

names = ["Albert","John","Richard","Henry","William"]
surnames = ["Goodman","Black","White","Green","Joneson"]
salaries = [500*random.randint(10,30) for _ in range(10)]

def generate_random_person(names, surnames, salaries):
    return {"name":random.sample(names,1)[0],
            "surname":random.sample(surnames,1)[0],
            "salary":random.sample(salaries,1)[0]}
def generate_people(k):
    return [generate_random_person(names, surnames, salaries) for _ in range(k)]

generate_random_person(names, surnames, salaries)

{'name': 'Richard', 'salary': 7500, 'surname': 'Joneson'}

df = pd.DataFrame(generate_people(50),columns=["name","surname","salary"])

df