03-08-2021 04:51 PM - edited 03-08-2021 04:54 PM
@eddy1990 wrote:
The data will be needed in a following Sub VI where I plan to take the inverse of the large 2d array.
That's going to be expensive and slow (assuming you mean "invert the matrix*", not just take 1/x of the elements).
*However, since your matrix contains NaN, the inverted matrix will be all NaN, so you need to clean it up first!
Can you give some background on what you are trying to achieve?
03-08-2021 06:39 PM - edited 03-08-2021 06:47 PM
Hi
The 2D array will contain only numeric values and NaN and Inf will be removed. I am working on solving linear equations of the form AX = B (nodal analysis). My A matrix is a 2D array of size 6900*6900. My B matrix is an array of size 6900*1. So I intend to perform inv(A)*B.
Since I want the whole 2D array, I was planning to use the New Data Value Reference. Will it improve the algorithm?
Thanks
03-08-2021 11:11 PM
A DVR will probably not improve anything here. Just don't wire any indicators to these gigantic arrays.
LabVIEW has direct tools to solve linear equations, so use these.