book-notes

• book link

I reccommend that you go over the jupyter notebooks / colab and play beyond.

Some questions that you can answer while doing the work are on each chapter, you can check the solutions here

Ipython

Ipython is a wonderful way to write new code interactively. I also usually have a text editor (VIM) and Ipython window open in paralell.

Command	Purpose
run scriptName.py	run script
?	shows docstrings
??	shows full code
whos	shows variables in memory

2. Numpy

Numpy arrays provide you with C-like speed for numerical operations

• What is the average height for US presidents?

3. Pandas

The Pandas library provides efficient storage and manipulation of dense type arrays in python.

You can define a pandas dataframe as:

• from a dict
• specify the index

Questions:

• What is the average population in the US if we only consider the states of California, Texas, New York, Florida, Illinois?

4. Matplotlib

Matplotlib is a plotting library that was made as a Matlab alternative

you can use a nicer style with axis style

plt.style.use('seaborn-white')
plt.style.use('seaborn-whitegrid') # add grid by default
plt.axis('tight')
plt.axis('equal')

you can also plot error bars

plt.errorbar(x, y, yerr=dy, fmt='.k');

As well as 3D data in 2D plots using

plt.contour, plt.contourf, and plt.imshow

Seaborn makes some nicer plots that matplotlib

import seaborn as sns sns.set()

5. Machine learning with Scikit-Learn

• supervised learning creates models of the data by fitting the data
• unsupervised learning is when you have no data labels