22 Jul 1991 ehg (hager) updated czero.f 49,50c49,52 < 20 IF ( N .LE. 1 ) RETURN < IF ( N .GT. 2 ) GOTO 30 --- > 20 IF ( N .GT. 1 ) GOTO 25 > Z(1) = -P(1) > RETURN > 25 IF ( N .GT. 2 ) GOTO 30 10 Sep 1991 ehg (hager) btrans, bvert bug fix fft, ffc fix one comment line From math.ufl.edu!hager Wed Nov 6 13:19:47 EST 1991 Date: Wed, 6 Nov 91 13:19:47 EST From: To: ehg@research.att.com Subject: more routines for napack Eric, Recently, I have added more routines to napack for processing complex matrices. I also made a few changes in earlier routines. A shell script for installing these programs in netlib/napack follows. Note that the index was also updated to reflect the new routines. In the shell script below, I assume that the index is still stored in the same directory as the other napack programs. Best wishes, Bill # From research!moss!rutgers!psuvax1!hager Thu Oct 1 22:57:15 1987 # Date: Thu, 1 Oct 87 20:35:02 EDT # From: rutgers!psuvax1!hager (William W. Hager) # Message-Id: <8710020035.AA16035@psuvax1.psu.edu> # To: dongarra@anl-mcs.arpa, moss!research!ehg # Subject: 2 replacement programs for napack # # Two subroutines (cahess and chess) in napack need an extra statement # to test if a term vanishes. I think that the easiest way to insert # this extra statement in the routines is to shell the following # script in the napack directory, overwriting the original subroutines # by the new ones. Sorry for the inconvenience. Bill. sed > cahess.f <<'#####' 's/^-//' -C -C ________________________________________________________ -C | | -C | BALANCE A COMPLEX MATRIX AND REDUCE TO HESSENBERG FORM | -C | | -C | INPUT: | From arpa!foxtrot.math.ufl.edu!hager Wed Mar 22 12:23:46 PST 1989 Date: Wed, 22 Mar 89 12:23:46 PST From: hager@foxtrot.math.ufl.edu Apparently-To: ehg@research.att.com >From hager@foxtrot.math.ufl.edu Wed Mar 22 12:21:41 1989 From: hager@foxtrot.math.ufl.edu Date: Wed, 22 Mar 89 15:18:23 -0500 To: na.grosse@na-net.stanford.edu Subject: oops oops, I sent you bidag, but it was subroutine sing that had the typo error in it. The updated version of 'sing.f' follows: From arpa!foxtrot.math.ufl.edu!hager Thu Mar 23 10:06:31 PST 1989 Date: Thu, 23 Mar 89 10:06:31 PST From: hager@foxtrot.math.ufl.edu Apparently-To: ehg@research.att.com >From hager@foxtrot.math.ufl.edu Thu Mar 23 10:05:57 1989 From: hager@foxtrot.math.ufl.edu Date: Thu, 23 Mar 89 13:02:59 -0500 To: na.grosse@na-net.stanford.edu Subject: cg again I just received a report from someone who received a compiler error for the new subroutine cg when he used the CRAY to compile it. Here is another copy of cg.f that takes care of the compiler error: Apr 92 stopit.f added save statement