Reading a summary in Python

Every block of the table answers a different question.

import numpy as np
import statsmodels.api as sm

rng = np.random.default_rng(2)
x = rng.normal(size=(50, 2))
y = 2 + 3 * x[:, 0] - 1.5 * x[:, 1] + rng.normal(0, 0.5, size=50)

model = sm.OLS(y, sm.add_constant(x)).fit()
text = model.summary().as_text()

print(len(text.splitlines()), "lines of summary")
print(round(model.aic, 2), round(model.bic, 2))
print(round(model.mse_resid, 4), model.df_resid, model.df_model)
print(model.tvalues.round(2))

How it works

The coefficient block is estimate, error, t and p.
R-squared and F test the model as a whole.
Durbin-Watson and Jarque-Bera flag assumption problems.

Keywords and builtins used here

as
len
print
round

The run, in numbers

Lines: 14
Characters to type: 446
Tokens: 163
Three-star pace: 105 tpm

At the three-star pace of 105 tokens a minute, this run takes about 93 seconds.

Type this snippet

Step 3 of 3 in Linear models, step 3 of 19 in Statistics with statsmodels.

← Previous Next →