An introduction to the data science workflow in Python

DC PyLadies | Thursday, 22 March 2018

by @angelaambroz

O hai!

These are the tutorial materials for the 22 March 2018 DC PyLadies meetup, An introduction to the data science workflow in Python.

These notebooks can be viewed and interacted with Binder: (note: this may take a while to load, as it installs all the libraries)

You can also view these notebooks here in GitHub (without interaction).

You can also download these notebooks and install everything on your local machine. To do so, choose a directory where you want to put this (path/of/your/choosing), go to that directory (cd path/of/your/choosing), clone these materials (git clone), install the Python packages with pip, and then launch the Jupyter notebooks (jupyter notebook).

cd path/of/your/choosing
git clone [email protected]:angelaambroz/2018_03_pyladies.git
pip install -r requirements.txt
jupyter notebook

Table of Contents

• 0_Introduction - Welcome to the tutorial! How to get data from files, from databases, and from APIs. An introduction to pandas.
• 1_EDA - Exploratory data analysis, using pandas and matplotlib. An introduction to visualizations.
• 2_StatsML - Fitting a linear regression using three different libraries: statsmodels, scikit-learn, and numpy. Comparing the results. Discussion of other libraries for machine learning and statistics.
• 3_Etc - Some recommended resources for learning more. Other things to learn.

Data sources

The Duke University dataset found in the two files in data/ is from github/Chrissymbeck. I found it via the Data is Plural newsletter.