solve_amat_beta¶
- sofia_redux.toolkit.resampling.solve_amat_beta(phi, data, weights)[source]¶
Convenience function returning matrices suitable for linear algebra.
Given independent variables \(\Phi\), data \(y\), and weights \(W\), returns matrices \(A\) and \(B\) where:
\[ \begin{align}\begin{aligned}A = \Phi W \Phi^T\\B = \Phi W y\end{aligned}\end{align} \]- Parameters:
- phinumpy.ndarray (n_terms, n_samples)
Polynomial terms of independent variables for each sample in the fit.
- datanumpy.ndarray (n_samples,)
Sample data values.
- weightsnumpy.ndarray (n_samples,)
Squared weights.
- Returns:
- A, Bnumpy.ndarray (n_terms, n_terms), numpy.ndarray (n_terms,)
The \(A\) and \(B\) terms described above.