Linear Model

stamox.regression.lm(data: Union[List, Tuple, pandas.core.frame.DataFrame, ArrayLike], formula = None, subset = None, weights = None, NA_action = 'drop', method = 'qr', dtype: dtype = <class 'jax.numpy.float64'>) -> OLSState

Fits a linear model using the given data and parameters.

Parameters:

Name	Type	Description	Default
data	Union[List, Tuple, DataFrame, ArrayLike]	The data to fit the linear model with.	required
formula	str	A formula for the linear model. Defaults to None.	None
subset	list	A list of indices to use as a subset of the data. Defaults to None.	None
weights	array-like	An array of weights to apply to the data. Defaults to None.	None
NA_action	str	The action to take when encountering missing values. Defaults to "drop".	'drop'
method	str	The method to use for fitting the linear model. Defaults to "qr".	'qr'
dtype	jnp.float——	The data type to use for the linear model. Defaults to jnp.float_.	<class 'jax.numpy.float64'>

Returns:

Type	Description
OLSState	The state of the fitted linear model.

Examples:

>>> from stamox.functions import lm
>>> import pandas as pd
>>> import numpy as np
>>> np.random.seed(42)
>>> X = np.random.uniform(size=(1000, 3))
>>> y = (
        3 * X[:, 0]
        + 2 * X[:, 1]
        - 7 * X[:, 2]
        + 1.
    )
>>> data = pd.DataFrame(
        np.concatenate([X, y.reshape((-1, 1))], axis=1),
        columns=["x1", "x2", "x3", "y"],
    )
>>> res = lm(data, "y ~ x1 + x2 + x3")
>>> res.coefs
Array([ 1., 3.,  2., -7.], dtype=float32)