marchenko_pastur_fit: Graph reconstruction algorithm.
In Fagan-Lab/constructnet: Network reconstruction from time series

Graph reconstruction algorithm.

marchenko_pastur_fit(
  TS,
  remove_largest = F,
  metric_distance = F,
  tol = 1e-15,
  threshold_type = "range",
  ...
)

`TS`	N * L matrix consisting of L observations from N variables.
`remove_largest`	If False, all the eigenvectors associated to the significant eigenvalues will be used to reconstruct the de-noised empirical correlation matrix. If True, the eigenvector associated to the largest eigenvalue is going to be excluded from the recontruction step.
`metric_distance`	If False, a signed graph is obtained. If True, the correlation is transformed by defining a metric distance between each pair of nodes
`tol`	avoid minor difference between python and r in case of calculating eigen values
`threshold_type`	Which thresholding function to use on the matrix of weights.
`...`	Arguments