Mod 5 matrices

Daniel Tafmizi

Dr. Friedman

Lis 4370

Module 5

Github: daniel.R/Work.R/LIS4370Rprog/Mod5.R at main · DanielDataGit/daniel.R (github.com)


The provided matrices were singular and non-square. For linear systems, finding an inverse of these types requires a variation of Singular Value Decomposition (SVD), which essentially breaks down a matrix into three parts (the rows, the columns, and its diagonal "time/instance order"). R's matlib library provides the function ginv(), which uses the Moore-Penrose inverse to create a generalized inverse that is useful in solving linear systems. The formula is A+=VD+UT, where V is an orthogonal row matrix, D+ is the transposed reciprocal of all non-zero values in the diagonal matrix, and UT is the transposed orthogonal column matrix. It is used to calculate an approximate solution to an "unsolvable" linear system. 

Comments

Popular posts from this blog

redditApiR

Mod 7 OOP

Description File