ggb_global_nopsd: Compute the Global GGB estimator without psd constraint
In jacobbien/ggb: Graph-Guided Banding of the Covariance Matrix

Uses the path BCD "closed form" of Yan and Bien (2015) as implemented in R package hsm.

1 2	ggb_global_nopsd(S, g, lambda = NULL, w = NULL, flmin = 0.01, nlam = 20, max_depths = NULL)

`S`	p-by-p sample covariance matrix
`g`	seed graph (not necessarily connected)
`lambda`	positive tuning parameter(s)
`w`	weights from b = 1 to diam(g). Default is sqrt( group size )
`flmin`	ratio of smallest to largest lambda value (ignored if `lambda` is nonnull)
`nlam`	number of lambda values (ignored if `lambda` is nonnull)
`max_depths`	maximal bandwidth considered. Should be a single non-negative integer. Default is NULL, which is equivalent to taking max_depths to be >= diameter(g). (This is referred to as M in the paper.)