[Nexus] how to read files from Eiger?

Pete Jemian jemian at anl.gov
Sun Jun 26 21:54:25 BST 2016


Thanks, all, for the various detailed advice.

My coworker and I have tried some of the suggestions and come to the 
conclusion that these compressed data files from the Eiger are not yet 
suitable for general distribution in places where other HDF5 files are 
handled routinely.

Assembling the special tools to read the data in these Eiger files is 
going to take much more patience to make work than is available now. 
Until the toolset becomes part of common distribution (we are using HDF 
1.8 libraries that are build into tools created by others), we will 
advise those who wish to analyze data with these compression formats to 
work with the instrument teams where the data was acquired.

Regards,
    Pete


On 6/24/2016 9:23 AM, Herbert J. Bernstein wrote:
> Dear Pete,
>
>   If you are on a system with good dlfnc support and are using a recent
> HDF5 library, then use of the plugin filters is transparent (just
> install the filters and define the HDF5_PLUGIN_PATH environment
> variable).  The serious problem is on a system or applications with an
> old hdf5 library (e.g. 1.6).  Then you need an application such as
> h5repack built against a recent hdf5 to repack the file with
> compressions that are known to the old application, but somewhere you
> need a new hdf5 installation to make this work.
>
>   This will make a good test case to document in detail on the
> hdrmx.medsbio.org <http://hdrmx.medsbio.org> website as a tutorial both
> for users and sysadmins.
>
>   Regards,
>     Herbert
>
> On Fri, Jun 24, 2016 at 8:45 AM, Jemian, Pete R. <jemian at anl.gov
> <mailto:jemian at anl.gov>> wrote:
>
>     Eugen
>
>     First off, I (representing the user who asked me about this) want to
>     read that file with the HDF5 plugin in IgorPro on a Mac.
>
>     Adding the compression library to this toolset is not trivial.  I
>     must persuade the 3rd party developer of that plugin to make this
>     addition.
>
>     I also want to read that file with the h5py package in python on
>     mac, Windows, and Linux.
>
>     We are on the leading edge here but our users have already begun to
>     struggle with reading these files.  Until the compression filters
>     are part of common installations, a converter tool is needed.
>
>     Pete
>
>
>
>
>     Sent from a device with a dreadfully small screen
>
>
>     -------- Original message --------
>     From: Eugen Wintersberger <eugen.wintersberger at desy.de
>     <mailto:eugen.wintersberger at desy.de>>
>     Date: 6/24/2016 3:05 AM (GMT-06:00)
>     To: "Jemian, Pete R." <jemian at anl.gov <mailto:jemian at anl.gov>>,
>     "Herbert J. Bernstein" <yayahjb at gmail.com
>     <mailto:yayahjb at gmail.com>>, Discussion forum for the NeXus data
>     format <nexus at nexusformat.org <mailto:nexus at nexusformat.org>>
>     Cc: Tobias.Richter at diamond.ac.uk
>     <mailto:Tobias.Richter at diamond.ac.uk>, jjakoncic at bnl.gov
>     <mailto:jjakoncic at bnl.gov>, andreas.foerster at dectris.com
>     <mailto:andreas.foerster at dectris.com>, rsweet at bnl.gov
>     <mailto:rsweet at bnl.gov>
>     Subject: Re: [Nexus] how to read files from Eiger?
>
>     Hi Pete,
>
>     On which Linux systems do you want to read the files. For Ubuntu
>     14.04 and Debian 8  we have packages available.
>
>     On 06/23/16 19:27, Pete Jemian wrote:
>>     Very helpful.  Thanks for the quick turnaround.
>>     Now I go off in a corner with an internet connection for quiet
>>     time ...
>>
>>     On 6/23/2016 12:25 PM, Herbert J. Bernstein wrote:
>>>     Dear Pete,
>>>
>>>       This is a solvable problem, but requires some preliminary
>>>     software installations.  First, be sure your HDF5 kit is
>>>     reasonably recent and has filter plugin support.  Then pickup and
>>>     install both the LZ4 and BitShuffle filters.  Now make sure
>>>     HDF5_PLUGIN_PATH is defined to point to the directory containing
>>>     the filter libraries.  Then h5dump should be able to handle the
>>>     Eiger files.
>     For the current files BittShuffle is not required. Most probably
>     required for latter releases of the detector?
>
>     regards
>        Eugen
>>>
>>>       Eugen is gathering the filters on github.  I am gathering other
>>>     software and links to test data and templates at
>>>     HDRMX.medsbio.org <http://HDRMX.medsbio.org>.  There is a link
>>>     for the hdrmx-bb list on the page.   Please post comments,
>>>     questions and suggestions there.
>>>
>>>       Regards,
>>>         Herbert
>>>
>>>     Sent from my iPad
>>>
>>>>     On Jun 23, 2016, at 12:48, Pete Jemian <prjemian at gmail.com>
>>>>     <mailto:prjemian at gmail.com> wrote:
>>>>
>>>>
>>>>     We have received a linked fileset from an Eiger 1M Anderson
>>>>     want to read it.
>>>>     Cannot seem to progress past knowing that there is data.
>>>>     Here is an excerpt from h5dump (version 1.8.16):
>>>>
>>>>            DATASET "data" {
>>>>               DATATYPE  H5T_STD_U32LE
>>>>               DATASPACE  SIMPLE { ( 1, 1065, 1030 ) / (
>>>>     H5S_UNLIMITED, 1065, 1030 ) }
>>>>               DATA {h5dump error: unable to print data
>>>>
>>>>               }
>>>>               ATTRIBUTE "image_nr_high" {
>>>>                  DATATYPE  H5T_STD_I32LE
>>>>                  DATASPACE  SCALAR
>>>>                  DATA {
>>>>                  (0): 1
>>>>                  }
>>>>               }
>>>>               ATTRIBUTE "image_nr_low" {
>>>>                  DATATYPE  H5T_STD_I32LE
>>>>                  DATASPACE  SCALAR
>>>>                  DATA {
>>>>                  (0): 1
>>>>                  }
>>>>               }
>>>>            }
>>>>
>>>>     It's the error: "h5dump error: unable to print data"
>>>>
>>>>     Is this the case where we have the wrong version of the HDF5
>>>>     libraries?
>>>>     Is there a tool to render such files compatible with the 1.8
>>>>     libraries?
>>>>     If so, where is that tool?
>>>>
>>>>     Pete
>>>>
>>>>     _______________________________________________
>>>>     NeXus mailing list
>>>>     NeXus at nexusformat.org <mailto:NeXus at nexusformat.org>
>>>>     http://lists.nexusformat.org/mailman/listinfo/nexus
>>>>
>>
>
>     --
>     Dr. Eugen Wintersberger
>
>     FS-EC
>     DESY
>     Notkestr. 85
>     D-22607 Hamburg
>     Germany
>
>     E-Mail: eugen.wintersberger at desy.de
>     <mailto:eugen.wintersberger at desy.de>
>     Telefon: +49-40-8998-1917 <tel:%2B49-40-8998-1917>
>
>

-- 
----------------------------------------------------------
Pete R. Jemian, Ph.D.                <jemian at anl.gov>
Beam line Controls and Data Acquisition, Group Leader
Advanced Photon Source,   Argonne National Laboratory
Argonne, IL  60439                   630 - 252 - 3189
-----------------------------------------------------------
    Education is the one thing for which people
       are willing to pay yet not receive.
-----------------------------------------------------------




More information about the NeXus mailing list