Loading ...
Sorry, an error occurred while loading the content.

using nmrShell -- need help

Expand Messages
  • Douglas Kojetin
    Hi All- I am trying to use nmrShell, beginning on one linux computer w/ 2 processors, but potentially expanding to multiple computers w/ 2 processors. I ve
    Message 1 of 4 , Oct 1, 2003
    • 0 Attachment
      Hi All-

      I am trying to use nmrShell, beginning on one linux computer w/ 2
      processors, but potentially expanding to multiple computers w/ 2
      processors.

      I've attempted to split up an HNCO processing script between the 2
      processors on the same computer (roadrunner), but it seems to only
      process one 'half' of what it is expected to ... and stalls.

      Can anyone provide any assistance? Sample scripts? Sample setup
      files? I'm not working with the 'runon' command under linux (as seen
      in the nmrShell help file for use under SGI systems; after a quick web
      search, I would have to install a linux cluster program and possibly
      recompile the kernel), but I know that running two processes at the
      same time should 'divide' the work between the processors, correct?
      Also, rsh is working.

      Many thanks in advance for the help.
      Doug


      misc output:

      ########################
      nmrShell:

      roadrunner 0:
      ------------------------------------------------------------------------
      -------------k
      | Processing XY Dimension.
      |
      | WaitFor: deleting file roadrunner.0.part1
      |
      | WaitFor: deleting file localhost.0.part1
      |
      | WaitFor: deleting file roadrunner.0.part2
      |
      | WaitFor: deleting file localhost.0.part2
      |
      | WaitFor: deleting file roadrunner.0.part3
      |
      | WaitFor: deleting file localhost.0.part3
      |
      | XYZ2Pipe Partition: Plane 1 to 34 of 68
      |
      | XY Plane 34 of 68

      localhost 0:
      ------------------------------------------------------------------------
      --------------k
      | Processing XY Dimension.
      |
      | WaitFor: deleting file roadrunner.0.part1
      |
      | WaitFor: deleting file localhost.0.part1
      |
      | WaitFor: deleting file roadrunner.0.part2
      |
      | WaitFor: deleting file localhost.0.part2
      |
      | WaitFor: deleting file roadrunner.0.part3
      |
      | WaitFor: deleting file localhost.0.part3
      |
      | XYZ2Pipe Partition: Plane 1 to 34 of 68
      |
      | XY Plane 34 of 68

      (this is where it stalls; at plane 34)

      ########################
      pipe.net:

      roadrunner 10
      localhost 10

      ########################
      other pipe.net:

      roadrunner 10
      roadrunner 10

      ########################
      ft_hnco.com:

      #!/bin/csh

      echo Processing XY Dimension.

      waitFor -par pipe.net -ext part1 -init -verb
      waitFor -par pipe.net -ext part2 -init -verb
      waitFor -par pipe.net -ext part3 -init -verb

      xyz2pipe -in cfids/hnco_092903_%03d.DAT -x -verb -par pipe.net \
      | nmrPipe -fn POLY -time \
      | nmrPipe -fn SP -off 0.45 -end 1.0 -pow 2 -c 0.5 \
      | nmrPipe -fn ZF -auto \
      | nmrPipe -fn FT \
      | nmrPipe -fn PS -p0 126.6 -p1 0.0 -di \
      | nmrPipe -fn EXT -x1 6.0ppm -xn 12.0ppm -sw \
      | pipe2xyz -out pdata/HNCO_%03d.ft12 -x -ov

      waitFor -par pipe.net -ext part1 -verb

      xyz2pipe -in pdata/HNCO_%03d.ft12 -y -verb -par pipe.net \
      | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 -c 1.0 \
      | nmrPipe -fn ZF -auto \
      | nmrPipe -fn FT \
      | nmrPipe -fn PS -p0 -90.0 -p1 180.0 -di \
      | pipe2xyz -out pdata/HNCO_%03d.ft13 -y -ov

      rm -rf pdata/HNCO_*.ft12

      echo Re-Processing Z Dimension With Linear Prediction

      waitFor -par pipe.net -ext part2 -verb

      xyz2pipe -in pdata/HNCO_%03d.ft13 -z -verb -par pipe.net \
      #| nmrPipe -fn LP -fb \
      | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 -c 0.5 \
      | nmrPipe -fn ZF -auto \
      | nmrPipe -fn FT \
      | nmrPipe -fn PS -p0 90.0 -p1 0.0 -di \
      | pipe2xyz -out pdata/HNCO_%03d.ft14 -z -ov

      rm -rf pdata/HNCO_*.ft13

      waitFor -par pipe.net -ext part3 -verb

      xyz2pipe -in pdata/HNCO_%03d.ft14 -x -verb -par pipe.net \
      | nmrPipe -fn POLY -auto \
      | pipe2xyz -out pdata/hnco_%03d.dat -x -ov

      rm -rf pdata/HNCO_*.ft14

      echo Done.
    • Frank Delaglio
      Many Greetings, When multiprocessing on a computer with a single hostname, you have to include -cpu arguments to distinguish between entries. The CPU number
      Message 2 of 4 , Oct 1, 2003
      • 0 Attachment
        Many Greetings,

        When multiprocessing on a computer with a single hostname,
        you have to include "-cpu" arguments to distinguish between
        entries. The CPU number is passed to your processing script
        as the first argument.

        xyz2pipe -in fid/test%03d.fid -x -verb -par pipe.net -cpu $1 \

        etc etc

        cheerful regards,

        big fd




        On Wed, 1 Oct 2003, Douglas Kojetin wrote:

        > Hi All-
        >
        > I am trying to use nmrShell, beginning on one linux computer w/ 2
        > processors, but potentially expanding to multiple computers w/ 2
        > processors.
        >
        > I've attempted to split up an HNCO processing script between the 2
        > processors on the same computer (roadrunner), but it seems to only
        > process one 'half' of what it is expected to ... and stalls.
        >
        > Can anyone provide any assistance? Sample scripts? Sample setup
        > files? I'm not working with the 'runon' command under linux (as seen
        > in the nmrShell help file for use under SGI systems; after a quick web
        > search, I would have to install a linux cluster program and possibly
        > recompile the kernel), but I know that running two processes at the
        > same time should 'divide' the work between the processors, correct?
        > Also, rsh is working.
        >
        > Many thanks in advance for the help.
        > Doug
        >
        >
        > misc output:
        >
        > ########################
        > nmrShell:
        >
        > roadrunner 0:
        > ------------------------------------------------------------------------
        > -------------k
        > | Processing XY Dimension.
        > |
        > | WaitFor: deleting file roadrunner.0.part1
        > |
        > | WaitFor: deleting file localhost.0.part1
        > |
        > | WaitFor: deleting file roadrunner.0.part2
        > |
        > | WaitFor: deleting file localhost.0.part2
        > |
        > | WaitFor: deleting file roadrunner.0.part3
        > |
        > | WaitFor: deleting file localhost.0.part3
        > |
        > | XYZ2Pipe Partition: Plane 1 to 34 of 68
        > |
        > | XY Plane 34 of 68
        >
        > localhost 0:
        > ------------------------------------------------------------------------
        > --------------k
        > | Processing XY Dimension.
        > |
        > | WaitFor: deleting file roadrunner.0.part1
        > |
        > | WaitFor: deleting file localhost.0.part1
        > |
        > | WaitFor: deleting file roadrunner.0.part2
        > |
        > | WaitFor: deleting file localhost.0.part2
        > |
        > | WaitFor: deleting file roadrunner.0.part3
        > |
        > | WaitFor: deleting file localhost.0.part3
        > |
        > | XYZ2Pipe Partition: Plane 1 to 34 of 68
        > |
        > | XY Plane 34 of 68
        >
        > (this is where it stalls; at plane 34)
        >
        > ########################
        > pipe.net:
        >
        > roadrunner 10
        > localhost 10
        >
        > ########################
        > other pipe.net:
        >
        > roadrunner 10
        > roadrunner 10
        >
        > ########################
        > ft_hnco.com:
        >
        > #!/bin/csh
        >
        > echo Processing XY Dimension.
        >
        > waitFor -par pipe.net -ext part1 -init -verb
        > waitFor -par pipe.net -ext part2 -init -verb
        > waitFor -par pipe.net -ext part3 -init -verb
        >
        > xyz2pipe -in cfids/hnco_092903_%03d.DAT -x -verb -par pipe.net \
        > | nmrPipe -fn POLY -time \
        > | nmrPipe -fn SP -off 0.45 -end 1.0 -pow 2 -c 0.5 \
        > | nmrPipe -fn ZF -auto \
        > | nmrPipe -fn FT \
        > | nmrPipe -fn PS -p0 126.6 -p1 0.0 -di \
        > | nmrPipe -fn EXT -x1 6.0ppm -xn 12.0ppm -sw \
        > | pipe2xyz -out pdata/HNCO_%03d.ft12 -x -ov
        >
        > waitFor -par pipe.net -ext part1 -verb
        >
        > xyz2pipe -in pdata/HNCO_%03d.ft12 -y -verb -par pipe.net \
        > | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 -c 1.0 \
        > | nmrPipe -fn ZF -auto \
        > | nmrPipe -fn FT \
        > | nmrPipe -fn PS -p0 -90.0 -p1 180.0 -di \
        > | pipe2xyz -out pdata/HNCO_%03d.ft13 -y -ov
        >
        > rm -rf pdata/HNCO_*.ft12
        >
        > echo Re-Processing Z Dimension With Linear Prediction
        >
        > waitFor -par pipe.net -ext part2 -verb
        >
        > xyz2pipe -in pdata/HNCO_%03d.ft13 -z -verb -par pipe.net \
        > #| nmrPipe -fn LP -fb \
        > | nmrPipe -fn SP -off 0.5 -end 0.98 -pow 2 -c 0.5 \
        > | nmrPipe -fn ZF -auto \
        > | nmrPipe -fn FT \
        > | nmrPipe -fn PS -p0 90.0 -p1 0.0 -di \
        > | pipe2xyz -out pdata/HNCO_%03d.ft14 -z -ov
        >
        > rm -rf pdata/HNCO_*.ft13
        >
        > waitFor -par pipe.net -ext part3 -verb
        >
        > xyz2pipe -in pdata/HNCO_%03d.ft14 -x -verb -par pipe.net \
        > | nmrPipe -fn POLY -auto \
        > | pipe2xyz -out pdata/hnco_%03d.dat -x -ov
        >
        > rm -rf pdata/HNCO_*.ft14
        >
        > echo Done.
        >
        >
        >
        >
        >
        >
        > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
        >
        >
        >
      • Atom Crystal
        Just for your information: A strange phenomenon has been daunting me for around three weeks and almost let me down (I even reinstalled the Linux system!)
        Message 3 of 4 , Oct 3, 2003
        • 0 Attachment
          Just for your information:

          A strange phenomenon has been daunting me for around
          three weeks and almost let me down (I even reinstalled
          the Linux system!) before suddenly I may have found
          the workaround.

          When processing a 3D spectrum, say, a CBCA(CO)NH
          acquired with x=HN, y=C, z=N, if I execute the fid ->
          ./data/test%03d.fid conversion script under the same
          directory a second time, overwriting existing
          test%03d.fid data files, and process as before,
          convert to NMRView format, check the spectrum
          projected along the y(=C) axis, i.e. view in HN-N
          axes(x=HN, y=N), I found the upper part of the
          projected spectrum shifted totally to right for about
          1ppm while the lower half kept in position, resulting
          in a projected spectrum where the peaks of the lower
          part corresponds well to those in a 2D HSQC while
          peaks in the upper part all shifted a certain ppm
          compared those in that 2D HSQC. This happens when all
          the converting/processing scripts are kept unchanged!

          This problem is reproducible on my Redhat Linux 9.0 on
          a Intel P4 for 3D Varian data. I have not tested
          Bruker data. The workaround is just to delete those
          directory containing test%03d.fid and test%03d.ft3
          before reprocessing.

          I'd wonder whether anyone met this before ( or maybe
          no one have ever tried to convert his fid to
          test%03d.fid two times?)

          Now the question is does var2pipe do some tricky
          things during conversion? Or it's just because that I
          set the wrong options for var2pipe, or I'm simply
          wrong in using var2pipe? Attached please find the
          script for conversion, generated by 'varian' and
          options set by "read parameters".

          Regards,

          Yufeng

          -----------------------------------

          #!/bin/csh

          var2pipe -in ./fid -noaswap -aqORD 1 \
          -xN 1024 -yN 96 -zN 56
          \
          -xT 512 -yT 48 -zT 28
          \
          -xMODE Complex -yMODE Complex -zMODE Rance-Kay
          \
          -xSW 6999.738 -ySW 12067.801 -zSW 1200.012
          \
          -xOBS 599.887 -yOBS 0.848 -zOBS 60.793
          \
          -xCAR 4.773 -yCAR 45.999 -zCAR 119.005
          \
          -xLAB HN -yLAB C -zLAB N
          \
          -ndim 3 -aq2D States
          \
          -out ./data/test%03d.fid -verb -ov

          sleep 5


          =====
          Yufeng Tong
          School of Medicine, E642A
          Dept. Physiol. & Biophys.
          CWRU, Cleveland, OH 44106, USA
          216-368-8654(O)
          --------------------------------------------------------
          Life is an everlasting game of weiqi (igo, baduk).

          __________________________________
          Do you Yahoo!?
          The New Yahoo! Shopping - with improved product search
          http://shopping.yahoo.com
        • Atom Crystal
          Thanks Big Frank, So I m simply wrong in using the in-place option. Below are quote of your email: The problem is not with the conversion command; rather it is
          Message 4 of 4 , Oct 7, 2003
          • 0 Attachment
            Thanks Big Frank,

            So I'm simply wrong in using the in-place option.

            Below are quote of your email:

            The problem is not with the conversion command; rather
            it is with the "-inPlace" processing. The second time
            you process the data, your directory "slp" contains
            "old" processed data, which gets mixed with newer data
            during the course of the processing scheme.

            When processing data in-place, it's best to start with
            an empty output directory to avoid this problem.


            --- Atom Crystal <crystalatom@...> wrote:
            > Just for your information:
            >
            > A strange phenomenon has been daunting me for around
            > three weeks and almost let me down (I even
            > reinstalled
            > the Linux system!) before suddenly I may have found
            > the workaround.
            >
            > When processing a 3D spectrum, say, a CBCA(CO)NH
            > acquired with x=HN, y=C, z=N, if I execute the fid
            > ->
            > ./data/test%03d.fid conversion script under the same
            > directory a second time, overwriting existing
            > test%03d.fid data files, and process as before,
            > convert to NMRView format, check the spectrum
            > projected along the y(=C) axis, i.e. view in HN-N
            > axes(x=HN, y=N), I found the upper part of the
            > projected spectrum shifted totally to right for
            > about
            > 1ppm while the lower half kept in position,
            > resulting
            > in a projected spectrum where the peaks of the lower
            > part corresponds well to those in a 2D HSQC while
            > peaks in the upper part all shifted a certain ppm
            > compared those in that 2D HSQC. This happens when
            > all
            > the converting/processing scripts are kept
            > unchanged!
            >
            > This problem is reproducible on my Redhat Linux 9.0
            > on
            > a Intel P4 for 3D Varian data. I have not tested
            > Bruker data. The workaround is just to delete those
            > directory containing test%03d.fid and test%03d.ft3
            > before reprocessing.
            >
            > I'd wonder whether anyone met this before ( or maybe
            > no one have ever tried to convert his fid to
            > test%03d.fid two times?)
            >
            > Now the question is does var2pipe do some tricky
            > things during conversion? Or it's just because that
            > I
            > set the wrong options for var2pipe, or I'm simply
            > wrong in using var2pipe? Attached please find the
            > script for conversion, generated by 'varian' and
            > options set by "read parameters".
            >
            > Regards,
            >
            > Yufeng
            >
            > -----------------------------------
            >
            > #!/bin/csh
            >
            > var2pipe -in ./fid -noaswap -aqORD 1 \
            > -xN 1024 -yN 96 -zN 56
            >
            > \
            > -xT 512 -yT 48 -zT 28
            >
            > \
            > -xMODE Complex -yMODE Complex -zMODE Rance-Kay
            >
            > \
            > -xSW 6999.738 -ySW 12067.801 -zSW 1200.012
            >
            > \
            > -xOBS 599.887 -yOBS 0.848 -zOBS 60.793
            >
            > \
            > -xCAR 4.773 -yCAR 45.999 -zCAR 119.005
            >
            > \
            > -xLAB HN -yLAB C -zLAB N
            >
            > \
            > -ndim 3 -aq2D States
            >
            > \
            > -out ./data/test%03d.fid -verb -ov
            >
            > sleep 5
            >
            >
            > =====
            > Yufeng Tong
            > School of Medicine, E642A
            > Dept. Physiol. & Biophys.
            > CWRU, Cleveland, OH 44106, USA
            > 216-368-8654(O)
            >
            --------------------------------------------------------
            > Life is an everlasting game of weiqi (igo, baduk).
            >
            > __________________________________
            > Do you Yahoo!?
            > The New Yahoo! Shopping - with improved product
            > search
            > http://shopping.yahoo.com
            >


            __________________________________
            Do you Yahoo!?
            The New Yahoo! Shopping - with improved product search
            http://shopping.yahoo.com
          Your message has been successfully submitted and would be delivered to recipients shortly.