From caa6605b434b78921d2665012b16b9680caadb7c Mon Sep 17 00:00:00 2001 From: Chulwoo Jung Date: Mon, 7 Aug 2017 00:01:04 -0400 Subject: [PATCH] Still tweaking memory saving routines in Lanczos --- lib/algorithms/iterative/ImplicitlyRestartedLanczosCJ.h | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-) diff --git a/lib/algorithms/iterative/ImplicitlyRestartedLanczosCJ.h b/lib/algorithms/iterative/ImplicitlyRestartedLanczosCJ.h index 5fc57774..800bd326 100644 --- a/lib/algorithms/iterative/ImplicitlyRestartedLanczosCJ.h +++ b/lib/algorithms/iterative/ImplicitlyRestartedLanczosCJ.h @@ -954,7 +954,7 @@ until convergence // Uses more temorary // Rotate0(Qt,evec,k1,k2,Nm); // Uses minimal temporary, possibly with less speed - Rotate(Qt,evec,k1-1,k2+1,Nm); + Rotate0(Qt,evec,k1-1,k2+1,Nm); // Try if Rotate() doesn't work // Rotate2(Qt,evec,k1,k2); t1=usecond()/1e6; @@ -1010,10 +1010,8 @@ until convergence evec[i] = B[Iconv[i]]; } #else - Rotate(Qt,evec,0,Nk,Nm); + Rotate0(Qt,evec,0,Nk,Nm); FinalCheck( Nk, eval,evec); -//exit(-1); -// ConvRotate2( Nk, Nm, Qt, evec, eval,eval2,Iconv,Nconv); #endif // Skip sorting, as it doubles the memory usage(!) and can be avoided by diagonalizing "right away"