Ordinary least squares in Python

The summary table is the point: coefficients with their uncertainty.

import numpy as np
import statsmodels.api as sm

rng = np.random.default_rng(0)
practice = np.arange(1, 61)
tpm = 70 + 0.7 * practice + rng.normal(0, 4, size=60)

X = sm.add_constant(practice)
model = sm.OLS(tpm, X).fit()

print(model.params.round(4))
print(model.bse.round(4), model.pvalues.round(6))
print(round(model.rsquared, 4), round(model.rsquared_adj, 4))
print(model.conf_int(alpha=0.05).round(3))

How it works

1. add_constant is required, or the model has no intercept.
2. fit returns a results object, not the model.
3. params, bse and pvalues are the numbers behind the table.

Keywords and builtins used here

• as
• print
• round

The run, in numbers

Lines

14

Characters to type

406

Tokens

139

Three-star pace

105 tpm

At the three-star pace of 105 tokens a minute, this run takes about 79 seconds.

Type this snippet

Step 1 of 3 in Linear models, step 1 of 19 in Statistics with statsmodels.

Next →