Reading is the secret! Careful reading of material provided will bless you now and in the long run.

• The idea of the Holy: 256 pages
• Journal Articles (I swear I have read this article almost 100 times)
• Optional References

Read 500 pages like this every day. That’s how knowledge works. It builds up, like compound interest. All of you can do it, but I guarantee not many of you will do it.

-Warren Buffett

Each project has a Readings section. See Project 1 for example. Let’s look at the readings and figure out what we need.

• Python for Data Science (P4DS): Data Visualization
• P4DS: Graphics for Communication
• P4DS: Markdown
• P4DS: 5.2 Filter rows with .query()

• Python
• VS Code
• Altair in VS Code

• VS Code, Python, and packages

In an interactive chunk run the following.

import sys
!{sys.executable} -m pip install numpy pandas scikit-learn altair

• altair_saver GitHub
• Altair Charts and NodeJS

Because VS Code fixed all the problems that I have historically had with pip. It just works now and is the method our students have been taught in CSE 110 and CSE 111.

• pip and conda

Let’s review the power of Python Interactive

• # %% in my .py script is much better than Jupyter notebooks (.ipynb).
  • If we hope to have our code work in a production environment then Jupyter is problematic.
  • Caching and code chunks are problematic
  • https://medium.com/@_orcaman/jupyter-notebook-is-the-cancer-of-ml-engineering-70b98685ee71

Setting up your script

A good data science .py script will have packages and data loaded at the top. Usually you have a few short commented sentences that descibe the script purpose.

# %%
# import pandas, altair, numpy
import pandas as pd
import altair as alt
import numpy as np

# %%
# load data
# handgrenade data https://github.com/byuidatascience/data4soils/blob/master/data-raw/cfbp_handgrenade/cfbp_handgrenade.csv

url = 'https://github.com/byuidatascience/data4soils/raw/master/data-raw/cfbp_handgrenade/cfbp_handgrenade.csv'

dat = pd.read_csv(url)

1. Encode the row and column to the axes.
2. Color the hmx points using the ‘goldorange’ color scheme.
3. Use mark_square() and make the square sizes 500.

1. Encode the x-axis as binned.
2. Encode the y-axis as counts.
3. Configure the title to a fontSize of 20.
4. Use properties to place the title.