fips.Independent

class fips.Independent

Bases: CorrelationModel

The identity correlation: each label is correlated with itself only.

By label, not by coordinates: two labels at the same place and time are still independent. Independent().with_std(std) is a diagonal covariance, and a nugget is a sum, model.with_std(a) + Independent().with_std(b).

Examples

>>> import pandas as pd
>>> from fips import Independent, Space
>>> space = Space({"co2": pd.Index(["a", "b"], name="site")})
>>> Independent().with_std(0.5).build(space).dense()
array([[0.25, 0.  ],
       [0.  , 0.25]])

Attributes

levels

No levels, since the identity reads none.

Methods

correlation(coords)

Return the identity matrix, one row per row of coords.

Inherited

From CorrelationModel: with_std