DFMResult.r_squared

DFMResult.r_squared()

Share of each series’ variance the common component explains.

On the standardized scale each column has unit variance, so the share is \(1 - \operatorname{Var}(e_i)\); its mean across series is variance_shares.sum().

Return type:

dict[str, float]

Returns:

A mapping from series name to its common-component R^2 on the standardized scale.

Example

>>> import numpy as np
>>> rng = np.random.default_rng(0)
>>> f = rng.standard_normal((150, 1))
>>> lam = np.array([[2.0, 1.0, 0.0]])
>>> panel = f @ lam + 0.5 * rng.standard_normal((150, 3))
>>> fit = DFM(panel, order=1, n_factors=1, series_names=("a", "b", "c")).fit()
>>> fit = fit.r_squared()
>>> list(fit), bool(fit["a"] > fit["b"] > fit["c"])
(['a', 'b', 'c'], True)
>>> bool(fit["a"] > 0.9), bool(fit["c"] < 0.1)
(True, True)