Errors when testing install of latest version

Errors when testing install of latest version

Postby gjtanoury » Tue Nov 04, 2014 2:18 pm

[Originally put this in the wrong section. Has been moved from General Discussion]

Hi Eric,
I just compiled and installed the Oct.26, 2014 version of the Linux x86_64 source code. I failed on three tests. I compiled with ifort v14.0.1. I cannot upload any files directly, as the forum seems to not allow the upload of any files with any extension.

Here's the output from the error files:
b.6.12.err
254 A: 10. LP ( 5)Cu 2 -- -- 135.0 90.0 -- -- -- --
254 N: 10. LP ( 5)Cu 2 -- -- 135.0 270.0 -- -- -- --
!! Error: 270.0 should be 90.0

b.8.4.2.err
1139 A: 5. BD ( 2) C 1- C 2 62.3 270.0 89.1 357.9 87.7 90.6 180.0 89.7
1139 N: 5. BD ( 2) C 1- C 2 62.3 270.0 89.1 182.1 87.7 90.6 180.0 89.7
!! Error: 182.1 should be 357.9

b.23.err
302 A: 0.0000 -0.0181 0.8610 0.0000 0.0000
298 N: 0.0000 -0.0000 0.0000 -0.0207 0.9894
!! Error: -0.0000 should be -0.0181
!! Error: 0.0000 should be 0.8610
!! Error: -0.0207 should be 0.0000
!! Error: 0.9894 should be 0.0000

If you would like to see my Make.config, let me know where I can send it, as it seems the forum does not allow me to upload Make.config

Jerry
gjtanoury
 
Posts: 33
Joined: Thu Feb 14, 2013 7:04 am

Re: Errors when testing install of latest version

Postby ericg » Wed Nov 05, 2014 11:17 pm

Jerry,

None of these, in fact, are of significant concern.

The first two are discrepancies in the hybrid directionality table. The analysis reports the locations of the density maximum for each hybrid. For b.6.12, LP(5) on Cu 2 is a d-type function with at least two (if not four) equivalent lobes. (135.0, 90.0) and (135.0, 270.0) correspond to the maxima of different lobes. Due to slight numerical differences in the calculations, the gfortran-compiled code yields one lobe, and ifort yields a different one. For b.8.4.2, BD(2) is a pi bond, with gfortran identifying one lobe of the C 1 2p orbital, and intel identifying the other lobe.

The third is a discrepancy in the character of a Rydberg orbital. Rydbergs are of low occupancy and are not particularly well determined by the NBO search procedure, so differences can be safely ignored.

Eric
ericg
 
Posts: 265
Joined: Sat Dec 29, 2012 9:31 am

Re: Errors when testing install of latest version

Postby gjtanoury » Mon Nov 10, 2014 12:53 pm

OK. Thank you.
gjtanoury
 
Posts: 33
Joined: Thu Feb 14, 2013 7:04 am

Re: Errors when testing install of latest version

Postby VCalvoP » Fri Jan 30, 2015 7:54 pm

Ho means Hodgkin; a great scientific; Ho uname gives this,
Linux Ho 3.11.10-100.fc18.x86_64 #1 SMP Mon Dec 2 20:28:38 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux
gamessNBO as I called it, compiled just fine and linked perfectly.
No error or warning issued. Was glad, but. Running nbo examples included just gave the usual gamess output.
On the other hand, gennbo script works just fine, in the NBO6 format. Therefore the problem is in compall I guess
Gimme a hint, thanks
VCalvoP
VCalvoP
 
Posts: 1
Joined: Sat Jan 24, 2015 8:14 pm

Re: Errors when testing install of latest version

Postby ericg » Sun Feb 01, 2015 11:28 am

Did you set NBO and NBOLIB in the lked script?

Eric
ericg
 
Posts: 265
Joined: Sat Dec 29, 2012 9:31 am


Return to NBO Installation

Who is online

Users browsing this forum: No registered users and 1 guest

cron