AMBER Archive (2007)

Subject: Re: AMBER: Could not find cntrl namelist

From: Robert Duke (rduke_at_email.unc.edu)
Date: Wed Jul 11 2007 - 08:07:36 CDT

Francesco -
Okay, this is purely a matter of nonprinting characters in the mdin file for
the MD RUN. Unfortunately, Ross and I were misled by the mdin,
"vac_min.in", which was for the minimization, and which was the only mdin
sent in the group of files to the list. SO I looked at the output of
eqmecycl_vac_md1_12Acut.out. This is the output that produces the "Could
not find cntrl namelist" message. If you look at the echo of the mdin
input, it looks fine here, and it is also clear we were looking at the wrong
mdin. However, if you do an "od -c eqmecycl_vac_md1_12Acut.out > od.dmp"
and look at od.dmp, you will see, with a little work, that there are two
nonprinting characters, octal 302 followed by octal 240, in front of every
mdin input line. Perhaps the mdin file was made on another operating
system, with a different native language, or some such? You need to use an
editor like vi on a compatible unix system, or edit the current one, getting
rid of the apparent blanks, which I would guess are some sort of unicode and
replacing them with true blanks. Then run the mdin you created through
od -c again, confirming that all the nonprinting characters that previously
showed up as octal numbers have disappeared. Most likely, this is all that
is going on. For my edification, I would be interested to know how you
generated this file; I have had trouble myself moving files from windows to
linux, but I think that was a matter of /r/n instead of /n at the end of the
lines (windows has a different line termination convention than unix). Have
fun with the dogs; if you fix the mdin, you really should be running.
Best Regards - Bob
----- Original Message -----
From: "Francesco Pietra" <chiendarret_at_yahoo.com>
To: <amber_at_scripps.edu>
Sent: Wednesday, July 11, 2007 6:14 AM
Subject: Re: AMBER: Could not find cntrl namelist

> Please, see below.
> --- "M. L. Dodson" <mldodson_at_houston.rr.com> wrote:
>
>> Francesco Pietra wrote:
>> > Hi Ross:
>> > I'll study your suggestions below, which arrived when I had decided to
>> simplify
>> > the problem by running equatorial methylcyclohexane, prepared for xleap
>> with
>> > antechamber. The same issue "Could not find cntrl namelost" arose.
>> >
>> > This time, contrary to all previous experience, I could quit the xleap
>> window
>> > without getting the "not saved" message. Contrary to my hope, it didn't
>> > anticipate the solution.
>> >
>> > I also run my eqeq molecule with different mdin, always getting the
>> > "Could
>> not
>> > ..." error message.
>> >
>> > Attached are all relevant files.
>> >
>> > Thanks
>> >
>> > francesco pietra
>> >
>> >
>>
>> This input file runs fine for me.
>
> Do you mean the preliminary minimization (which runs fine for me, too), or
> the
> subsequent MD? (which is where I get the error message).
>
>> I think your sander compilation
>> must be bogus somehow. I'm using sander compiled with g95.
>
> I installed Amber9 serial (to begin with) on debian linux amd64 etch.
> Compilers
> used were ifort_x86_64 (that I indicated) and gcc (that was used
> automatically;
> now that machine is switched off, can't verify the version, though is
> surely an
> updated one; I never installed g95). Sander failed to compile static (as I
> was
> indicated to do), then I switched to dynamic, it compiled and all tests
> passed
> without errors.
>
> I had previously installed openmpi and libnuma, though parallelization was
> not
> carried out (planned with all intel compilers).
>
> Can you suggest an extra test for sander serial?
>
> Thanks
>
> francesco pietra
>
>>
>> Bud Dodson
>>
>> PS, where did you get the mdinfo file? I got one that differs
>> only in the number of steps, 272 vs. 274 (and everything else on
>> the NSTEP data line, but differs only in the last few decimal
>> places).
>>
>> >
>> >
>> > --- Ross Walker <ross_at_rosswalker.co.uk> wrote:
>> >
>> >> Hi Francesco,
>> >>
>> >>> What I fear at this point is related to what I posted
>> >>> yesterday about a remind
>> >>> by David Case that antechamber treats single residues:
>> >> The issue is that the only possible things that could be generating
>> >> this
>> >> error message is either there is something wrong with the mdin file,
>> >> the
>> >> incorrect file was specified with -i or there is some issue with read
>> >> permissions of the mdin file. At this point in the code the prmtop
>> >> file
>> >> hasn't even been opened so it cannot be related to a problem with
>> >> that.
>> >>
>> >> Could you please attach your mdin file to an email (as an actual
>> attachment)
>> >> direct from the copy on the machine on which you are having problems.
>> >> If
>> you
>> >> have to ftp it to another machine to email it then make sure you
>> >> select
>> >> binary format for the transfer. Then we can take a look and see if
>> >> there
>> is
>> >> something weird going on.
>> >>
>> >> Another option you could try as well is to create a brand new file on
>> >> the
>> >> machine on which you are running sander and explicitly type in, on
>> >> that
>> >> machine in vi, everything you want in the input file. That way you
>> >> work
>> >> around any kind of weird corruption and/or strange control characters
>> >> that
>> >> may have made it into the old file during the copying / cut / pasting
>> >> etc.
>> >>
>> >> Seeing the exact input file as sander sees it on your machine will
>> >> really
>> >> help.
>> >>
>> >> All the best
>> >> Ross
>> >>
>> >> /\
>> >> \/
>> >> |\oss Walker
>> >>
>> >> | HPC Consultant and Staff Scientist |
>> >> | San Diego Supercomputer Center |
>> >> | Tel: +1 858 822 0854 | EMail:- ross_at_rosswalker.co.uk |
>> >> | http://www.rosswalker.co.uk | PGP Key available on request |
>> >>
>> >> Note: Electronic Mail is not secure, has no guarantee of delivery, may
>> >> not
>> >> be read every day, and should not be used for urgent or sensitive
>> >> issues.
>> >>
>> >>
>> >> -----------------------------------------------------------------------
>> >> The AMBER Mail Reflector
>> >> To post, send mail to amber_at_scripps.edu
>> >> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>> >>
>> >
>> >
>> >
>> >
>> >
>>
> ____________________________________________________________________________________Ready
>> for the edge of your seat?
>> > Check out tonight's top picks on Yahoo! TV.
>> > http://tv.yahoo.com/
>>
>> --
>> M. L. Dodson
>> Email: mldodson-at-houston-dot-rr-dot-com
>> Phone: eight_three_two-five_63-386_one
>> -----------------------------------------------------------------------
>> The AMBER Mail Reflector
>> To post, send mail to amber_at_scripps.edu
>> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>>
>
>
>
>
> ____________________________________________________________________________________
> Building a website is a piece of cake. Yahoo! Small Business gives you all
> the tools to get online.
> http://smallbusiness.yahoo.com/webhosting
> -----------------------------------------------------------------------
> The AMBER Mail Reflector
> To post, send mail to amber_at_scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu
>

-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber_at_scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo_at_scripps.edu