Data Visualization in Python

Univariate Plots

We will be introducing plotting and code from 3 modules: matplotlib, seaborn and pandas. As we go forth, you may ask the question, which one should I learn? Chris Moffitt has the following advice.

A pathway to learning (Chris Moffit)

1. Learn the basic matplotlib terminology, specifically what is a Figure and an Axes .
2. Always use the object-oriented interface. Get in the habit of using it from the start of your analysis. (not really getting into this, but basically, don't use the Matlab form I'll show at the end if you don't have to)
3. Start your visualizations with basic pandas plotting.
4. Use seaborn for the more complex statistical visualizations.
5. Use matplotlib to customize the pandas or seaborn visualization.
   
   

mtcars.plot.hist(y = 'mpg');
plt.show()
# mtcars.plot(y = 'mpg', kind = 'hist')
#mtcars['mpg'].plot(kind = 'hist')

mtcars['cyl'].value_counts().plot.bar();
plt.show()

mtcars['mpg'].plot( kind = 'density');
plt.show()

ax = sns.distplot(mtcars['mpg'], kde=False);
plt.show()

sns.countplot(data = mtcars, x = 'cyl');
plt.show()

diamonds = pd.read_csv('data/diamonds.csv.gz')
ordered_colors = ['E','F','G','H','I','J']
sns.catplot(data = diamonds, x = 'color', kind = 'count', color = 'blue');
plt.show()

sns.distplot(mtcars['mpg'], hist=False);
plt.show()