[Pw_forum] Pw_forum Digest, Vol 21, Issue 81

ding yi ding.y1ding at gmail.com
Sat Mar 28 13:03:03 CET 2009 

Subject: Re: Line 325 in bands_FS.f90
I think that there is a mistake in the bands_FS.f90. In the bands_FS.f90,
line 310," backspace(5) "
should be replace by "
! back nlines+1 positions (number of eigenvalues lines plus one blank line) !
do k=1,nlines+1
backspace(5)
enddo !
"
The bands_FS counts the lines of the spin-polarized calculations, while for
the non-spin-polarised case it forgets to counts.


Yi Ding, Ph.D Candidate Department of Physics, Tsinghua University Beijing
100084, P.R.China

2009/3/27  <pw_forum-request at pwscf.org>:
> Send Pw_forum mailing list submissions to
>        pw_forum at pwscf.org
>
> To subscribe or unsubscribe via the World Wide Web, visit
>        http://www.democritos.it/mailman/listinfo/pw_forum
> or, via email, send a message with subject or body 'help' to
>        pw_forum-request at pwscf.org
>
> You can reach the person managing the list at
>        pw_forum-owner at pwscf.org
>
> When replying, please edit your Subject line so it is more specific
> than "Re: Contents of Pw_forum digest..."
>
>
> Today's Topics:
>
>   1. Line 325 in bands_FS.f90 (Paul M. Grant)
>   2. Re: Line 325 in bands_FS.f90 (Axel Kohlmeyer)
>   3. Re: Line 325 in bands_FS.f90 (Eyvaz Isaev)
>   4. Re: Line 325 in bands_FS.f90 (Eyvaz Isaev)
>   5. Re: Line 325 in bands_FS.f90 (Paul M. Grant)
>   6. Re: Line 325 in bands_FS.f90 (Eyvaz Isaev)
>
>
> ----------------------------------------------------------------------
>
> Message: 1
> Date: Thu, 26 Mar 2009 12:34:19 -0700
> From: "Paul M. Grant" <w2agz at pacbell.net>
> Subject: [Pw_forum] Line 325 in bands_FS.f90
> To: "'PWSCF Forum'" <pw_forum at pwscf.org>
> Message-ID: <016001c9ae49$dcaada80$96008f80$@net>
> Content-Type: text/plain; charset="us-ascii"
>
> While running bands_FS.x (espresso-4.0 distro) on the output of an PW nscf
> non-spin-polarized job, I get the following error:
>
> At line 325 of file bands_FS.f90
> Fortran runtime error: Bad real number in item 5 of list input
>
> This only happens for my LSDA=F runs, spin-polarized are OK.  (BTW, the
> LSDA=F run for Ni in example 08 does work.  The only difference is that
> ex08 deals with only 8 bands and 1 formated line, whilst my task computes 20
> taking up 3 formated lines.)  Line 325 is in the LSDA=F portion of the code
> follows a commented-out read command line with explicit formatting.  I can't
> see any problem with the output of my nscf run.  Before I start messing
> around with the bands_FS.f90 source, has anyone else run into this issue
> (Eyvaz?)?  Is it possible there is a "line counting" error in the LSDA=F
> portion when the band energy listing exceeds one line?  Sorry if I've made
> some stupid blunder.
>
> Paul M. Grant, PhD
> Physicist and Science Writer
> Principal, W2AGZ Technologies
> Visiting Scholar, Applied Physics, Stanford (2005-2008)
> EPRI Science Fellow (Retired)
> IBM Research Staff Member Emeritus
> w2agz at pacbell.net
> http://www.w2agz.com <http://www.w2agz.com/>
>
>
>
>
> -------------- next part --------------
> An HTML attachment was scrubbed...
> URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090326/6d182aed/attachment-0001.htm
>
> ------------------------------
>
> Message: 2
> Date: Thu, 26 Mar 2009 16:00:07 -0400
> From: Axel Kohlmeyer <akohlmey at cmm.chem.upenn.edu>
> Subject: Re: [Pw_forum] Line 325 in bands_FS.f90
> To: PWSCF Forum <pw_forum at pwscf.org>
> Message-ID: <1238097607.3704.119.camel at zero>
> Content-Type: text/plain; charset="us-ascii"
>
> On Thu, 2009-03-26 at 12:34 -0700, Paul M. Grant wrote:
>> While running bands_FS.x (espresso-4.0 distro) on the output of an PW
>> nscf non-spin-polarized job, I get the following error:
>>
>> At line 325 of file bands_FS.f90
>>
>> Fortran runtime error: Bad real number in item 5 of list input
>
> paul,
>
> please check your input carefully with hexdump to make
> sure you have no non-ascii characters in there. with some
> fonts it is hard to tell. i'm attaching a wrapper script that
> i use myself, which is little bit easier to read (at least for
> me, as it emulates the hexdumper display that i was
> using back in the days when i was hacking 6502 processors
> which almost broke the 1MHz barrier...). i miss begin able
> to fix bugs in running programs by replacing goto opcodes
> with NOPs... ;-)
>
> cheers,
>   axel.
>
>
>>
>> This only happens for my LSDA=F runs, spin-polarized are OK.  (BTW,
>> the LSDA=F run for Ni in example 08 does work.  The only difference is
>> that
>>
>> ex08 deals with only 8 bands and 1 formated line, whilst my task
>> computes 20 taking up 3 formated lines.)  Line 325 is in the LSDA=F
>> portion of the code follows a commented-out read command line with
>> explicit formatting.  I can't see any problem with the output of my
>> nscf run.  Before I start messing around with the bands_FS.f90 source,
>> has anyone else run into this issue (Eyvaz?)?  Is it possible there is
>> a "line counting" error in the LSDA=F portion when the band energy
>> listing exceeds one line?  Sorry if I've made some stupid blunder.
>>
>> Paul M. Grant, PhD
>>
>> Physicist and Science Writer
>>
>> Principal, W2AGZ Technologies
>>
>> Visiting Scholar, Applied Physics, Stanford (2005-2008)
>>
>> EPRI Science Fellow (Retired)
>>
>> IBM Research Staff Member Emeritus
>>
>> w2agz at pacbell.net
>>
>> http://www.w2agz.com
>>
>>
>>
>>
>>
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>
> --
> =======================================================================
> Axel Kohlmeyer   akohlmey at cmm.chem.upenn.edu   http://www.cmm.upenn.edu
>   Center for Molecular Modeling   --   University of Pennsylvania
> Department of Chemistry, 231 S.34th Street, Philadelphia, PA 19104-6323
> tel: 1-215-898-1582,  fax: 1-215-573-6233,  office-tel: 1-215-898-5425
> =======================================================================
> If you make something idiot-proof, the universe creates a better idiot.
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: hexview
> Type: application/x-shellscript
> Size: 231 bytes
> Desc: not available
> Url : http://www.democritos.it/pipermail/pw_forum/attachments/20090326/e2878900/attachment-0001.bin
>
> ------------------------------
>
> Message: 3
> Date: Thu, 26 Mar 2009 13:50:46 -0700 (PDT)
> From: Eyvaz Isaev <eyvaz_isaev at yahoo.com>
> Subject: Re: [Pw_forum] Line 325 in bands_FS.f90
> To: PWSCF Forum <pw_forum at pwscf.org>
> Message-ID: <593553.85654.qm at web65705.mail.ac4.yahoo.com>
> Content-Type: text/plain; charset=us-ascii
>
>
> Dear Paul,
>
> Thanks reporting the error. I will look at why it happens.
>
> Bests,
> Eyvaz.
>
>
> --- On Thu, 3/26/09, Paul M. Grant <w2agz at pacbell.net> wrote:
>
>> From: Paul M. Grant <w2agz at pacbell.net>
>> Subject: [Pw_forum] Line 325 in bands_FS.f90
>> To: "'PWSCF Forum'" <pw_forum at pwscf.org>
>> Date: Thursday, March 26, 2009, 10:34 PM
>> While running bands_FS.x (espresso-4.0 distro) on the output
>> of an PW nscf
>> non-spin-polarized job, I get the following error:
>>
>> At line 325 of file bands_FS.f90
>> Fortran runtime error: Bad real number in item 5 of list
>> input
>>
>> This only happens for my LSDA=F runs, spin-polarized are
>> OK.  (BTW, the
>> LSDA=F run for Ni in example 08 does work.  The only
>> difference is that
>> ex08 deals with only 8 bands and 1 formated line, whilst my
>> task computes 20
>> taking up 3 formated lines.)  Line 325 is in the LSDA=F
>> portion of the code
>> follows a commented-out read command line with explicit
>> formatting.  I can't
>> see any problem with the output of my nscf run.  Before I
>> start messing
>> around with the bands_FS.f90 source, has anyone else run
>> into this issue
>> (Eyvaz?)?  Is it possible there is a "line
>> counting" error in the LSDA=F
>> portion when the band energy listing exceeds one line?
>> Sorry if I've made
>> some stupid blunder.
>>
>> Paul M. Grant, PhD
>> Physicist and Science Writer
>> Principal, W2AGZ Technologies
>> Visiting Scholar, Applied Physics, Stanford (2005-2008)
>> EPRI Science Fellow (Retired)
>> IBM Research Staff Member Emeritus
>> w2agz at pacbell.net
>> http://www.w2agz.com <http://www.w2agz.com/>
>>
>>
>>
>>
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
>
>
> ------------------------------
>
> Message: 4
> Date: Thu, 26 Mar 2009 14:14:31 -0700 (PDT)
> From: Eyvaz Isaev <eyvaz_isaev at yahoo.com>
> Subject: Re: [Pw_forum] Line 325 in bands_FS.f90
> To: PWSCF Forum <pw_forum at pwscf.org>
> Message-ID: <620097.49786.qm at web65711.mail.ac4.yahoo.com>
> Content-Type: text/plain; charset=us-ascii
>
>
> Dear Paul,
>
> I have just tested the code provided in 4.0.4, but with the output of 3.2. Both, Non-Spin polarized, and Spin-Polarized results worked well for Ni.
>
> And you can see from the line 325 that there is no limitation for bands numbers.
>
> Nevertheless, I will try the code with 4.0.4 output.
> By the way, it might be much helpful, if you send me your output file for which you have troubles.
>
> Bests,
> Eyvaz.
>
>
> --- On Thu, 3/26/09, Paul M. Grant <w2agz at pacbell.net> wrote:
>
>> From: Paul M. Grant <w2agz at pacbell.net>
>> Subject: [Pw_forum] Line 325 in bands_FS.f90
>> To: "'PWSCF Forum'" <pw_forum at pwscf.org>
>> Date: Thursday, March 26, 2009, 10:34 PMI
>> While running bands_FS.x (espresso-4.0 distro) on the output
>> of an PW nscf
>> non-spin-polarized job, I get the following error:
>>
>> At line 325 of file bands_FS.f90
>> Fortran runtime error: Bad real number in item 5 of list
>> input
>>
>> This only happens for my LSDA=F runs, spin-polarized are
>> OK.  (BTW, the
>> LSDA=F run for Ni in example 08 does work.  The only
>> difference is that
>> ex08 deals with only 8 bands and 1 formated line, whilst my
>> task computes 20
>> taking up 3 formated lines.)  Line 325 is in the LSDA=F
>> portion of the code
>> follows a commented-out read command line with explicit
>> formatting.  I can't
>> see any problem with the output of my nscf run.  Before I
>> start messing
>> around with the bands_FS.f90 source, has anyone else run
>> into this issue
>> (Eyvaz?)?  Is it possible there is a "line
>> counting" error in the LSDA=F
>> portion when the band energy listing exceeds one line?
>> Sorry if I've made
>> some stupid blunder.
>>
>> Paul M. Grant, PhD
>> Physicist and Science Writer
>> Principal, W2AGZ Technologies
>> Visiting Scholar, Applied Physics, Stanford (2005-2008)
>> EPRI Science Fellow (Retired)
>> IBM Research Staff Member Emeritus
>> w2agz at pacbell.net
>> http://www.w2agz.com <http://www.w2agz.com/>
>>
>>
>>
>>
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
>
>
> ------------------------------
>
> Message: 5
> Date: Thu, 26 Mar 2009 16:14:31 -0700
> From: "Paul M. Grant" <w2agz at pacbell.net>
> Subject: Re: [Pw_forum] Line 325 in bands_FS.f90
> To: <eyvaz_isaev at yahoo.com>,    "'PWSCF Forum'" <pw_forum at pwscf.org>
> Message-ID: <017b01c9ae68$9f10a5f0$dd31f1d0$@net>
> Content-Type: text/plain;       charset="iso-8859-1"
>
> Axel and Eyvaz, thanks guys for the quick feedback.
>
> Eyvaz, I'll send the files under separate cover...let me do a few more
> checks first so I don't wind up wasting your time.  However, I am simply
> using standard pw.x nscf output. Perhaps some bad characters slipped in, but
> I don't see how. Incidentally, a number of post-processing tools like
> bands_FS.f90 parse the pw.x text output files.  I wonder if future releases
> of PWscf will preserve the present layout.  Already I've added a number of
> personal flags to electrons.f90 to help "grep" out stuff I want, but I don't
> think any of these are causing the present problem, as they're in the scf
> output, not nscf.
>
> Axel, thanks for the hexdump script. I use the distro supplied tool
> occasionally, but this is a nice wrapper.  It reminds me of very useful byte
> editors we had in TSO/CMS, and fv.exe, an internal IBM utility we had in the
> original DOS.  You're going to find this hilarious.  I downloaded
> hexview.dat onto my Windoze laptop for a quick look, and then "samba-ed" to
> my Linux boxes.  Of course, the newline rat-tracks were left behind, and the
> bash interpreter didn't like that "one bit."
>
> Regards, -Paul
>
>
>
> Paul M. Grant, PhD
> Physicist and Science Writer
> Principal, W2AGZ Technologies
> Visiting Scholar, Applied Physics, Stanford (2005-2008)
> EPRI Science Fellow (Retired)
> IBM Research Staff Member Emeritus
> w2agz at pacbell.net
> http://www.w2agz.com
> ?
> ?
>
>
> -----Original Message-----
> From: pw_forum-bounces at pwscf.org [mailto:pw_forum-bounces at pwscf.org] On
> Behalf Of Eyvaz Isaev
> Sent: Thursday, March 26, 2009 2:15 PM
> To: PWSCF Forum
> Subject: Re: [Pw_forum] Line 325 in bands_FS.f90
>
>
> Dear Paul,
>
> I have just tested the code provided in 4.0.4, but with the output of 3.2.
> Both, Non-Spin polarized, and Spin-Polarized results worked well for Ni.
>
> And you can see from the line 325 that there is no limitation for bands
> numbers.
>
> Nevertheless, I will try the code with 4.0.4 output.
> By the way, it might be much helpful, if you send me your output file for
> which you have troubles.
>
> Bests,
> Eyvaz.
>
>
> --- On Thu, 3/26/09, Paul M. Grant <w2agz at pacbell.net> wrote:
>
>> From: Paul M. Grant <w2agz at pacbell.net>
>> Subject: [Pw_forum] Line 325 in bands_FS.f90
>> To: "'PWSCF Forum'" <pw_forum at pwscf.org>
>> Date: Thursday, March 26, 2009, 10:34 PMI
>> While running bands_FS.x (espresso-4.0 distro) on the output
>> of an PW nscf
>> non-spin-polarized job, I get the following error:
>>
>> At line 325 of file bands_FS.f90
>> Fortran runtime error: Bad real number in item 5 of list
>> input
>>
>> This only happens for my LSDA=F runs, spin-polarized are
>> OK.  (BTW, the
>> LSDA=F run for Ni in example 08 does work.  The only
>> difference is that
>> ex08 deals with only 8 bands and 1 formated line, whilst my
>> task computes 20
>> taking up 3 formated lines.)  Line 325 is in the LSDA=F
>> portion of the code
>> follows a commented-out read command line with explicit
>> formatting.  I can't
>> see any problem with the output of my nscf run.  Before I
>> start messing
>> around with the bands_FS.f90 source, has anyone else run
>> into this issue
>> (Eyvaz?)?  Is it possible there is a "line
>> counting" error in the LSDA=F
>> portion when the band energy listing exceeds one line?
>> Sorry if I've made
>> some stupid blunder.
>>
>> Paul M. Grant, PhD
>> Physicist and Science Writer
>> Principal, W2AGZ Technologies
>> Visiting Scholar, Applied Physics, Stanford (2005-2008)
>> EPRI Science Fellow (Retired)
>> IBM Research Staff Member Emeritus
>> w2agz at pacbell.net
>> http://www.w2agz.com <http://www.w2agz.com/>
>>
>>
>>
>>
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
> ------------------------------
>
> Message: 6
> Date: Thu, 26 Mar 2009 16:29:37 -0700 (PDT)
> From: Eyvaz Isaev <eyvaz_isaev at yahoo.com>
> Subject: Re: [Pw_forum] Line 325 in bands_FS.f90
> To: PWSCF Forum <pw_forum at pwscf.org>
> Message-ID: <347405.96383.qm at web65708.mail.ac4.yahoo.com>
> Content-Type: text/plain; charset=iso-8859-1
>
>
> Paul,
>
> OK, I'll wait for your output file. Anyway, I decided to put a new check for bands number, as one guy put more bands number that allowed (for n_last in  input_FS) and then claimed that the code does not work.
>
> Bests,
> Eyvaz.
>
>
>
> --- On Fri, 3/27/09, Paul M. Grant <w2agz at pacbell.net> wrote:
>
>> From: Paul M. Grant <w2agz at pacbell.net>
>> Subject: Re: [Pw_forum] Line 325 in bands_FS.f90
>> To: eyvaz_isaev at yahoo.com, "'PWSCF Forum'" <pw_forum at pwscf.org>
>> Date: Friday, March 27, 2009, 2:14 AM
>> Axel and Eyvaz, thanks guys for the quick feedback.
>>
>> Eyvaz, I'll send the files under separate cover...let
>> me do a few more
>> checks first so I don't wind up wasting your time.
>> However, I am simply
>> using standard pw.x nscf output. Perhaps some bad
>> characters slipped in, but
>> I don't see how. Incidentally, a number of
>> post-processing tools like
>> bands_FS.f90 parse the pw.x text output files.  I wonder if
>> future releases
>> of PWscf will preserve the present layout.  Already
>> I've added a number of
>> personal flags to electrons.f90 to help "grep"
>> out stuff I want, but I don't
>> think any of these are causing the present problem, as
>> they're in the scf
>> output, not nscf.
>>
>> Axel, thanks for the hexdump script. I use the distro
>> supplied tool
>> occasionally, but this is a nice wrapper.  It reminds me of
>> very useful byte
>> editors we had in TSO/CMS, and fv.exe, an internal IBM
>> utility we had in the
>> original DOS.  You're going to find this hilarious.  I
>> downloaded
>> hexview.dat onto my Windoze laptop for a quick look, and
>> then "samba-ed" to
>> my Linux boxes.  Of course, the newline rat-tracks were
>> left behind, and the
>> bash interpreter didn't like that "one bit."
>>
>> Regards, -Paul
>>
>>
>>
>> Paul M. Grant, PhD
>> Physicist and Science Writer
>> Principal, W2AGZ Technologies
>> Visiting Scholar, Applied Physics, Stanford (2005-2008)
>> EPRI Science Fellow (Retired)
>> IBM Research Staff Member Emeritus
>> w2agz at pacbell.net
>> http://www.w2agz.com
>> ?
>> ?
>>
>>
>> -----Original Message-----
>> From: pw_forum-bounces at pwscf.org
>> [mailto:pw_forum-bounces at pwscf.org] On
>> Behalf Of Eyvaz Isaev
>> Sent: Thursday, March 26, 2009 2:15 PM
>> To: PWSCF Forum
>> Subject: Re: [Pw_forum] Line 325 in bands_FS.f90
>>
>>
>> Dear Paul,
>>
>> I have just tested the code provided in 4.0.4, but with the
>> output of 3.2.
>> Both, Non-Spin polarized, and Spin-Polarized results worked
>> well for Ni.
>>
>> And you can see from the line 325 that there is no
>> limitation for bands
>> numbers.
>>
>> Nevertheless, I will try the code with 4.0.4 output.
>> By the way, it might be much helpful, if you send me your
>> output file for
>> which you have troubles.
>>
>> Bests,
>> Eyvaz.
>>
>>
>> --- On Thu, 3/26/09, Paul M. Grant
>> <w2agz at pacbell.net> wrote:
>>
>> > From: Paul M. Grant <w2agz at pacbell.net>
>> > Subject: [Pw_forum] Line 325 in bands_FS.f90
>> > To: "'PWSCF Forum'"
>> <pw_forum at pwscf.org>
>> > Date: Thursday, March 26, 2009, 10:34 PMI
>> > While running bands_FS.x (espresso-4.0 distro) on the
>> output
>> > of an PW nscf
>> > non-spin-polarized job, I get the following error:
>> >
>> > At line 325 of file bands_FS.f90
>> > Fortran runtime error: Bad real number in item 5 of
>> list
>> > input
>> >
>> > This only happens for my LSDA=F runs, spin-polarized
>> are
>> > OK.  (BTW, the
>> > LSDA=F run for Ni in example 08 does work.  The only
>> > difference is that
>> > ex08 deals with only 8 bands and 1 formated line,
>> whilst my
>> > task computes 20
>> > taking up 3 formated lines.)  Line 325 is in the
>> LSDA=F
>> > portion of the code
>> > follows a commented-out read command line with
>> explicit
>> > formatting.  I can't
>> > see any problem with the output of my nscf run.
>> Before I
>> > start messing
>> > around with the bands_FS.f90 source, has anyone else
>> run
>> > into this issue
>> > (Eyvaz?)?  Is it possible there is a "line
>> > counting" error in the LSDA=F
>> > portion when the band energy listing exceeds one line?
>>
>> > Sorry if I've made
>> > some stupid blunder.
>> >
>> > Paul M. Grant, PhD
>> > Physicist and Science Writer
>> > Principal, W2AGZ Technologies
>> > Visiting Scholar, Applied Physics, Stanford
>> (2005-2008)
>> > EPRI Science Fellow (Retired)
>> > IBM Research Staff Member Emeritus
>> > w2agz at pacbell.net
>> > http://www.w2agz.com <http://www.w2agz.com/>
>> >
>> >
>> >
>> >
>> > _______________________________________________
>> > Pw_forum mailing list
>> > Pw_forum at pwscf.org
>> > http://www.democritos.it/mailman/listinfo/pw_forum
>>
>>
>>
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>>
>> _______________________________________________
>> Pw_forum mailing list
>> Pw_forum at pwscf.org
>> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
>
>
>
> ------------------------------
>
> _______________________________________________
> Pw_forum mailing list
> Pw_forum at pwscf.org
> http://www.democritos.it/mailman/listinfo/pw_forum
>
>
> End of Pw_forum Digest, Vol 21, Issue 81
> ****************************************
>

More information about the Pw_forum mailing list