bend | R Documentation |
Bending a symmetric non-positive-definite matrix to positive-definite, using weighted or unweighted methods.
bend( inmat, wtmat, reciprocal = FALSE, max.iter = 10000, small.positive = 1e-04, method = "hj" )
inmat |
: The |
wtmat |
: The weight |
reciprocal |
: If |
max.iter |
: Maximum number of iterations. If no input is provided, default = 10000. |
small.positive |
: Eigenvalues smaller than this value are replaced with this value. If no input is provided, default = 0.0001. |
method |
: |
bent : The bent matrix
.
init.ev : Eigenvalues of the initial (inmat
) matrix.
final.ev : Eigenvalues of the bent
matrix.
min.dev : min(bent - inmat)
.
max.dev : max(bent - inmat)
.
loc.min.dev : Location (indices) of min.dev
element.
loc.max.dev : Location (indices) of max.dev
element.
ave.dev : Average deviation (bent - inmat
) of the upper triangle elements (excluding diagonal elements for correlation matrices).
AAD : Average absolute deviation of the upper triangle elements (excluding diagonal elements for correlation matrices) of bent
and inmat
.
Cor : Correlation between the upper triangle elements (excluding diagonal elements for correlation matrices) of bent
and inmat
.
RMSD : Root of mean squared deviation of the upper triangle elements (excluding diagonal elements for correlation matrices) of bent
and inmat
.
w_gt_0 : Number of weight elements greater than 0, in the upper triangle of wtmat
(for weighted bending).
wAAD : Weighted AAD
(for weighted bending).
wCor : Weighted Cor
(for weighted bending).
wRMSD : Weighted RMSD
(for weighted bending).
# Test data V = matrix(nrow=5, ncol=5, c( # matrix to be bent 100, 95, 80, 40, 40, 95, 100, 95, 80, 40, 80, 95, 100, 95, 80, 40, 80, 95, 100, 95, 40, 40, 80, 95, 100)) W = matrix(nrow=5, ncol=5, c( # matrix of weights 1000, 500, 20, 50, 200, 500, 1000, 500, 5, 50, 20, 500, 1000, 20, 20, 50, 5, 20, 1000, 200, 200, 50, 20, 200, 1000)) # Example 1: Unweighted bending bend(V) ## The default method (Jojani et al. 2003) is used. # Example 2: Weighted bending using reciprocal of the weighting factors bend(inmat=V, wtmat=W, reciprocal=TRUE) # Example 3: Bending with fixed elements ## Assume we want to keep V[1:2, 1:2] constant. W2 = W; W2[1:2, 1:2] = 0 bend(inmat=V, wtmat=W2, reciprocal=TRUE) # Example 4: Bending a correlation matrix V2 = cov2cor(V) bend(V2, W, reciprocal=TRUE) # Example 5: Bending using the method of Schaeffer (2014) bend(inmat=V, method="lrs") # Example 6: Bending a correlation matrix using the method of Schaeffer (2014) bend(V2, method="lrs") # Example 7: Bending the same correlation matrix using a weighted development of Schaeffer (2014) bend(V2, W, reciprocal=TRUE, method="lrs") # Example 8: Bending a covariance matrix using a weighted development of Schaeffer (2014) bend(V, W, reciprocal=TRUE, method="lrs")
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.