statsmodels.stats.correlation_tools.cov_nearest_factor_homog¶

statsmodels.stats.correlation_tools.
cov_nearest_factor_homog
(cov, rank)[source]¶ Approximate an arbitrary square matrix with a factorstructured matrix of the form k*I + XX’.
 Parameters
 covarraylike
The input array, must be square but need not be positive semidefinite
 rankpositive integer
The rank of the fitted factor structure
 Returns
 A FactoredPSDMatrix instance containing the fitted matrix
Notes
This routine is useful if one has an estimated covariance matrix that is not SPD, and the ultimate goal is to estimate the inverse, square root, or inverse square root of the true covariance matrix. The factor structure allows these tasks to be performed without constructing any n x n matrices.
The calculations use the fact that if k is known, then X can be determined from the eigendecomposition of cov  k*I, which can in turn be easily obtained form the eigendecomposition of cov. Thus the problem can be reduced to a 1dimensional search for k that does not require repeated eigendecompositions.
If the input matrix is sparse, then cov  k*I is also sparse, so the eigendecomposition can be done effciciently using sparse routines.
The onedimensional search for the optimal value of k is not convex, so a local minimum could be obtained.
Examples
Hard thresholding a covariance matrix may result in a matrix that is not positive semidefinite. We can approximate a hard thresholded covariance matrix with a PSD matrix as follows:
>>> import numpy as np >>> np.random.seed(1234) >>> b = 1.5  np.random.rand(10, 1) >>> x = np.random.randn(100,1).dot(b.T) + np.random.randn(100,10) >>> cov = np.cov(x) >>> cov = cov * (np.abs(cov) >= 0.3) >>> rslt = cov_nearest_factor_homog(cov, 3)