Graphics

• import matplotlib.pyplot as plt
• setup graph area: fig, ax = subplots(figsize=(w,h))
  • resize figure: fig.set_size_inches(w,h)
  • multiple plots: fig, axes = subplots(nrows=2,ncols=3,figsize=(15,5))
    • affect individual plot: axes[r,c].plot(x,y)
• plot: ax.plot(x,y);
  • line plot is default
  • scatterplot: ax.plot(x, y, 'o'); or ax.scatter(x, y, marker='o');
  • labels (methods) : ax.set_xlabel(), ax.set_ylabel(), ax.set_title()
• save: fig.savefig("Figure.png", dpi=400)
• 3-dimensional data
  • contour map: ax.contour(x,y,z)
  • heatmap: ax.imshow()

Additional Graphical and Numerical Summaries

• plot values in df columns: ax.plot(df['col1'], df['col2'], 'o');
• use df.plot() to call the variables by name and auto populate axes
  • i.e. ax = Auto.plot.scatter('horsepower', 'mpg')
  • multiple plots in one chart: Auto.plot.scatter('horsepower', 'mpg', ax=axes[1]);
• update column type: pd.Series()
  • i.e. Auto.cylinders = pd.Series(Auto.cylinders, dtype='category')
• boxplot: df.boxplot()
• histogram: df.hist()
• scatterplot matrix: pd.plotting.scatter_matrix(df,figsize=(16,16));
• numerical summary: df.describe() or df[col].describe()
  residual scatter plot:

ax = subplots(figsize=(8,8))[1]
ax.scatter(results.fittedvalues, results.resid)
ax.set_xlabel('Fitted value')
ax.set_ylabel('Residual')
ax.axhline(0, c='k', ls='--');

leverage scatter plot:

infl = results.get_influence()
ax = subplots(figsize=(8,8))[1]
ax.scatter(np.arange(X.shape[0]), infl.hat_matrix_diag)
ax.set_xlabel('Index')
ax.set_ylabel('Leverage')

you can plot a single variable super fast with df.plot(y='col')