Conversation
| const Real& A = Sk.Get(i,0); | ||
| const Real& B = S.Get(i,0); | ||
| Output( "Sk = ", A, "SVD: S_k", B); | ||
| auto eigenvalue_error = A - B; |
There was a problem hiding this comment.
I presume that this should be Abs(A-B) instead of A-B since you are testing its value being greater than a threshold later. (Also, the line is inconsistently indented and using snake_case for a variable. Also, why use A and B for scalars; that is a bit confusing when most of the library uses Householder notation for matrices and vectors.)
| auto eigenvalue_error = A - B; | ||
| auto left_svec_error = Norm(Abs(Uk( ALL, IR(i))) - Abs(U( ALL, IR(i)))); | ||
| auto right_svec_error = Norm(Abs(Vk( ALL, IR(i))) - Abs(V( ALL, IR(i)))); | ||
| if( eigenvalue_error > limits<Real>::Epsilon()){ |
There was a problem hiding this comment.
I think you mean El::limits::Epsilon<Real>(); std::numeric_limits<Real> does not support many of Elemental's datatypes (e.g., BigFloat). The syntax is slightly different and this is the first problem breaking your Travis build.
It would also be a good idea to use a more principled failure condition, such as max(m,n) || A ||_F eps
If this option is not set, the communication will land on the Hydrogen-default stream. This _could_ be a separate stream from the data compute stream, but in the LBANN use case, it generally will be the compute stream.
If this option is not set, the communication will land on the Hydrogen-default stream. This _could_ be a separate stream from the data compute stream, but in the LBANN use case, it generally will be the compute stream.
4 participants
I'm issuing this pull request to track progress on adding a Lanczos TSVD implementation to Elemental.
This may be slightly premature, so I am submitting a PR to track any work that is needed on this.
Added @andreasnoack and @jiahao. If you guys don't mind reviewing it would be much appreciated.