md2sInner: md2sInner

In peterjbachman/md2s: Multi-Dataset Multidimensional Scaling

md2sInner

Description

Determines which method of solution to solve the given system. In md2s, user selects either singular value decomposition or pseudoinverse. This function computes each method of solving. First, it runs conditionals to check whether pseudoinverse is appropriate. Then, it runs the method for solution, solves least squares from the solution, and then computes correlation coefficients between normed data sets. It then runs log likelihood for OLS parameters. In essance, if the system is square, solve the system with singular value decomposition. If the system is not square, compute and solve the system with pseudoinverse.

Usage

md2sInner(
  X0,
  y0,
  X.c.0 = NULL,
  X.X.0 = NULL,
  X.y.0 = NULL,
  init0 = "svd",
  tol0 = 1e-06
)

Arguments

X0	double centered matrix of X from md2s passed into md2sinner
y0	double centered matrix of X from md2s passed into md2sinner
X.c.0	empty matrix of covariates in shared space
X.X.0	empty matrix of covariates in X space, alone not shared with that of Y
X.y.0	empty matrix of covariates in Y space, alone not shared with that of X
init0	option set to singular value decomposition, although pseudoinverse is run if svd is not logical in md2sinner
tol0	tolerance param

Value

a list with the following

z	See md2s documentation
z.X	See md2s documentation
z.y	See md2s documentation
w.Xs	See md2s documentation
w.ys	See md2s documentation
beta.z	Coefficient from solution.See md2s documentation
proportionX	Proportion of X in y. See md2s documentation

Author(s)

Peter Bachman [e-mail], Patrick Edwards [e-mail], and Zion Little [e-mail]

peterjbachman/md2s documentation built on May 13, 2022, 8:14 p.m.