DOCSLIB.ORG

The Hierarchical Data Format (HDF): a Foundation for Sustainable Data and Software

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
The Hierarchical Data Format (HDF): a Foundation for Sustainable Data and Software The Hierarchical Data Format (HDF): A Foundation for Sustainable Data and Software Ted Habermann, The HDF Group; Andrew Collette, IMPACT, University of Colorado; Steve Vincena, BAPSF, University of California; Jay Jay Billings, Oak Ridge National Lab; Matt GerrinG, Diamond Light Source; Pierre de Buyl, KU Leuven; Konrad Hinson, Centre de Biophysique Moléculaire, Orléans (France); Mark Könnecke, Paul Scherrer Institut; Werner BenGer, Airborne HydroMappinG GmbH and Louisiana State University; Filipe RNC Maia, Uppsala University; Suren Byna, Lawrence Berkeley National Laborato- ry Abstract When combined with community specific conventions and application-specific semantics, the Hierarchical Data Format (HDF) provides a robust and reliable foundation that many scien- tific communities rely on for interoperability and high-performance storage of metadata and data. This foundation addresses many technical obstacles related to sustainable data and soft- ware, critical steps towards sustainable science. Building on successes across the communities that are using, or could use, HDF5 successfully, requires significant community engagement and development efforts that are outside of the scope of traditional software development projects in research and non-profit organizations like The HDF Group. Identifying individuals and organi- zations that are already doing the right things and bringing them together to share best practic- es for funding and executing these activities is an important step towards our shared goal of sustainable science. Introduction Sustainable science depends on intricate relationships between data and the software tools that access and analyze them. Data must be preserved in well-documented, self-describing formats accessible on multiple platforms using many programming languages. In addition, the formats must include metadata required for using and understanding the data long after they are collected and processed. In the short-term, sustainable formats and commercial, open-source, and community tools must support specific data and analysis needs of multiple scientific communities in order to achieve the usage and support required for sustaining maintenance and development of new capabilities. In practice, this is achieved through the development of conventions that map community science data types to computer science data types in the file. These conventions ensure interoperability by providing consistency in the data layer and isolating science users from the details of the underlying storage organization and format. The Hierarchical Data Format (HDF5) addresses these needs and forms a foundation for data storage and access in many scientific communities. The flexibility of the format is achieved through simplicity. HDF5 is a container format that includes three fundamental objects: da- tasets, attributes, and groups. Datasets are multi-dimensional arrays of atomic or compound computer science datatypes. They provide the flexibility required for storage of diverse scien- tific data objects while attributes allow annotation of these objects with metadata required for use and understanding. Groups provide organization of datasets and attributes into associa- tions that define scientific data objects. Combined with community specific conventions and application-specific semantics, HDF provides a robust and reliable foundation that scientific communities rely on for interoperabil- ity and high-performance storage of metadata and data. The communities described here have overcome many obstacles to sustainable data and software. Finding effective mechanisms for sharing and building on their experiences is an exciting next step towards sustainable science. Scientific Facilities Many important scientific problems require investments in equipment or facilities that must be amortized across user communities. Scientists from all over the world travel to facilities to run experiments or use observations collected at/by these facilities. Multiple members of di- verse science teams must be able to freely share original observations and derived results without worrying about platforms, operating systems and other computer details. The HDF5 format is well suited to these demands and is being used successfully in many large facilities. The NASA dust accelerator facility at the University of Colorado's Institute for Modeling Plasma, Atmospheres and Cosmic Dust (impact.colorado.edu) is entirely standardized on HDF5. The data product is an HDF5 container following conventions that represent particle im- pact events from the accelerator. The "science data type" is a sequence of dust-particle events, each of which has an associated set of time series representing the experimental data, and as- sociated metadata. HDF5 files produced according to this scheme may be read using facility- provided IDL scripts, but users are free to use any HDF5-aware analysis environment (IDL, Matlab, Python, etc.) to explore the files. The use of standard HDF5 attributes and datasets means that "free-form" analysis in these environments is possible, without relying on the facili- ty scripts or special-purpose readers. The Basic Plasma Science Facility (BAPSF, http://plasma.physics.ucla.edu) is a national user facility (supported by DOE and NSF) dedicated to experiments addressing fundamental ques- tions in many sub-fields of plasma physics. Measurements are made with user-selectable probes in the 20 m long device. Time series from these probes are digitized and stored in HDF5 using facility-specific conventions. Additional data and metadata are stored in the file to docu- ment digitizer settings, the run-time state of other programmable devices associated with the experiment, as well as diagnostic information for the plasma device itself. The Diamond Light Source (DLS, http://www.diamond.ac.uk) is the United Kingdom’s syn- chrotron, another large scientific facility. DLS uses the ‘NeXus’ conventions with HDF5 for most of their experimental data bringing many benefits to users: HDF5 is built in to the next genera- tion of faster detectors, visualization and slicing of data, analysis of data larger than memory, users can create/test tools on one slice of data and process the whole stack, saving state or re- sults to a HDF5 file. In addition a wide range of tools are used to convert to and from HDF5. The final example involves modeling an entire facility in HDF5. The Reactor Analyzer is being developed at Oak Ridge (Billings et al., 2014) for reading, writing and viewing simulation results that uses HDF5 to handle all file processing. The system works by mapping the physical struc- tures of light-water and sodium-cooled fast reactors (reactor vessel, fuel pins, etc.) and the nu- clear plant (heat exchangers, pipes, etc.) to simple representative classes. Instances of these classes are then written to HDF5 data types, creating a completely consistent, transferable, hi- erarchical digital description of the nuclear reactor that was simulated. Models and Tools In other scientific communities, models are fundamental tools that increase understanding. HDF5 supports input and output from models as well as visualization and analysis of results. For example, in molecular chemistry and physics, molecular coordinates constitute the basic information for simulation software. In addition to spatial coordinates, many additional fields are critical for complete understanding: velocities, masses, charges, etc. The scientific data items are mostly scalars and vectors organized in arrays with elements for specific atoms. Be- yond the organization of the storage, an important challenge remains in the interpretation of the data in different contexts (biomolecular systems, simple fluids, etc.). H5MD (http://nongnu.org/h5md/), has recently been proposed as an HDF5-based file format for molecular data (P. de Buyl et al., 2014). It addresses data organization and definition within a HDF5 file and is already being used as a basis for further specialization with K. Hinsen propos- ing a H5MD module for his MOlecular SimulAtion Interchange Conventions (MOSAIC, 2014). The "F5" model (http://www.fiberbundle.net) utilizes the concept of fiber bundles, gluing differential geometry, topology and geometric algebra to cast many spatio-temporal data cate- gories into a hierarchy of five semantic levels. This concept fits perfectly into the capabilities of HDF5 to provide a concrete implementation of this abstract model. The F5 model has been used to efficiently store data sets from medical imaging, computational fluid dynamics, molecu- lar dynamics and geosciences. The F5 model allows to seamlessly integrate both raster and vec- tor graphics, as well as complex data categories such as curvilinear multiblock, unstructured n- dimensional meshes and arbitrarily attributed fragmented point clouds. H5hut (http://www-vis.lbl.gov/Research/H5hut/) provides an API for parallel HDF5 encapsu- lating several data models of particle-based simulations. Particle simulations run on supercom- puters with hundreds of thousands of cores in order to study plasma and accelerator physics. These applications simulate billions to trillions of particles and produce tens of gbytes to hun- dreds of tbytes of data. High-performance I/O, and analysis and visualization tools for these da- ta can be challenging, especially for domain scientists. The H5hut library provides an efficient implementation of I/O and data management operations hiding the complexity of parallel HDF5 and is simple to use. Other frameworks
Load more

Recommended publications
  • Parallel Data Analysis Directly on Scientific File Formats
    Parallel Data Analysis Directly on Scientific File Formats Spyros Blanas 1 Kesheng Wu 2 Surendra Byna 2 Bin Dong 2 Arie Shoshani 2 1 The Ohio State University 2 Lawrence Berkeley National Laboratory [email protected] {kwu, sbyna, dbin, ashosani}@lbl.gov ABSTRACT and physics, produce massive amounts of data. The size of Scientific experiments and large-scale simulations produce these datasets typically ranges from hundreds of gigabytes massive amounts of data. Many of these scientific datasets to tens of petabytes. For example, the Intergovernmen- are arrays, and are stored in file formats such as HDF5 tal Panel on Climate Change (IPCC) multi-model CMIP-5 and NetCDF. Although scientific data management systems, archive, which is used for the AR-5 report [22], contains over petabytes such as SciDB, are designed to manipulate arrays, there are 10 of climate model data. Scientific experiments, challenges in integrating these systems into existing analy- such as the LHC experiment routinely store many gigabytes sis workflows. Major barriers include the expensive task of of data per second for future analysis. As the resolution of preparing and loading data before querying, and convert- scientific data is increasing rapidly due to novel measure- ing the final results to a format that is understood by the ment techniques for experimental data and computational existing post-processing and visualization tools. As a con- advances for simulation data, the data volume is expected sequence, integrating a data management system into an to grow even further in the near future. existing scientific data analysis workflow is time-consuming Scientific data are often stored in data formats that sup- and requires extensive user involvement.
    [Show full text]
  • A Metadata Based Approach for Supporting Subsetting Queries Over Parallel HDF5 Datasets
    A Metadata Based Approach For Supporting Subsetting Queries Over Parallel HDF5 Datasets Thesis Presented in Partial Fulfillment of the Requirements for the Degree Master of Science in the Graduate School of The Ohio State University By Vignesh Santhanagopalan, B.S. Graduate Program in Computer Science and Engineering The Ohio State University 2011 Thesis Committee: Dr. Gagan Agrawal, Advisor Dr. Radu Teodorescu ABSTRACT A key challenge in scientific data management is to manage the data as the data sizes are growing at a very rapid speed. Scientific datasets are typically stored using low-level formats which store the data as binary. This makes specification of the data and processing very hard. Also, as the volume of data is huge, parallel configurations must be used to process the data to enable efficient access. We have developed a data virtualization approach for supporting subsetting queries on scientific datasets stored in native format. The data is stored in Hierarchical Data Format (HDF5) which is one of the popular formats for storing scientific data. Our system supports SQL queries using the Select, From and Where clauses. We support queries based on the dimensions of the dataset and also queries which are based on the dimensions and attributes (which provide extra information about the dataset) of the dataset. In order to support the different types of queries, we have the pre-processing and post-processing modules. We also parallelize the selection queries involving the dimensions and the attributes. Our system offers the following advantages. We provide SQL like abstraction for specifying subsets of interest which is a powerful mechanism.
    [Show full text]
  • HUDDL for Description and Archive of Hydrographic Binary Data
    University of New Hampshire University of New Hampshire Scholars' Repository Center for Coastal and Ocean Mapping Center for Coastal and Ocean Mapping 4-2014 HUDDL for description and archive of hydrographic binary data Giuseppe Masetti University of New Hampshire, Durham, [email protected] Brian R. Calder University of New Hampshire, Durham, [email protected] Follow this and additional works at: https://scholars.unh.edu/ccom Recommended Citation G. Masetti and Calder, Brian R., “Huddl for description and archive of hydrographic binary data”, Canadian Hydrographic Conference 2014. St. John's, NL, Canada, 2014. This Conference Proceeding is brought to you for free and open access by the Center for Coastal and Ocean Mapping at University of New Hampshire Scholars' Repository. It has been accepted for inclusion in Center for Coastal and Ocean Mapping by an authorized administrator of University of New Hampshire Scholars' Repository. For more information, please contact [email protected]. Canadian Hydrographic Conference April 14-17, 2014 St. John's N&L HUDDL for description and archive of hydrographic binary data Giuseppe Masetti and Brian Calder Center for Coastal and Ocean Mapping & Joint Hydrographic Center – University of New Hampshire (USA) Abstract Many of the attempts to introduce a universal hydrographic binary data format have failed or have been only partially successful. In essence, this is because such formats either have to simplify the data to such an extent that they only support the lowest common subset of all the formats covered, or they attempt to be a superset of all formats and quickly become cumbersome. Neither choice works well in practice.
    [Show full text]
  • Towards Interactive, Reproducible Analytics at Scale on HPC Systems
    Towards Interactive, Reproducible Analytics at Scale on HPC Systems Shreyas Cholia Lindsey Heagy Matthew Henderson Lawrence Berkeley National Laboratory University Of California, Berkeley Lawrence Berkeley National Laboratory Berkeley, USA Berkeley, USA Berkeley, USA [email protected] [email protected] [email protected] Drew Paine Jon Hays Ludovico Bianchi Lawrence Berkeley National Laboratory University Of California, Berkeley Lawrence Berkeley National Laboratory Berkeley, USA Berkeley, USA Berkeley, USA [email protected] [email protected] [email protected] Devarshi Ghoshal Fernando Perez´ Lavanya Ramakrishnan Lawrence Berkeley National Laboratory University Of California, Berkeley Lawrence Berkeley National Laboratory Berkeley, USA Berkeley, USA Berkeley, USA [email protected] [email protected] [email protected] Abstract—The growth in scientific data volumes has resulted analytics at scale. Our work addresses these gaps. There are in a need to scale up processing and analysis pipelines using High three key components driving our work: Performance Computing (HPC) systems. These workflows need interactive, reproducible analytics at scale. The Jupyter platform • Reproducible Analytics: Reproducibility is a key com- provides core capabilities for interactivity but was not designed ponent of the scientific workflow. We wish to to capture for HPC systems. In this paper, we outline our efforts that the entire software stack that goes into a scientific analy- bring together core technologies based on the Jupyter Platform to create interactive, reproducible analytics at scale on HPC sis, along with the code for the analysis itself, so that this systems. Our work is grounded in a real world science use case can then be re-run anywhere. In particular it is important - applying geophysical simulations and inversions for imaging to be able reproduce workflows on HPC systems, against the subsurface.
    [Show full text]
  • Parsing Hierarchical Data Format (HDF) Files Karl Nyberg Grebyn Corporation P
    Parsing Hierarchical Data Format (HDF) Files Karl Nyberg Grebyn Corporation P. O. Box 47 Sterling, VA 20167-0047 703-406-4161 [email protected] ABSTRACT This paper presents a description of the creation of a library to parse Hierarchical Data Format (HDF) Files in Ada. It describes a “work in progress” with discussion of current performance, limitations and future plans. Categories and Subject Descriptors D.2.2 [Design Tools and Techniques]: Software Libraries; E.2 [Data Storage Representations]: Object Representation; I.2.10 [Vision and Scene Understanding]: Representations, data structures, and transforms. General Terms Algorithms, Performance, Experimentation, Data Formats. Keywords Ada, HDF. 1. INTRODUCTION Large quantities of scientific data are published each year by NASA. These data are often accompanied by metadata files that describe the contents of individual files of the data. One example of this data is the ASTER (Advanced Spaceborne Thermal Emission and Reflection Radiometer) [1]. Each file of data (consisting of satellite imagery) in HDF (Hierarchical Data Format) [2] is accompanied by a metadata file in XML (Extensible Markup Language) [3], encoded according to a published DTD (Document Type Description) that indicates the components and types of data in the metadata file. Each ASTER data file consists of an image of the satellite as it passes over the earth [4]. Information on the location of the data collected as the satellite passes is contained in the metadata file. Over time, multiple images of the same location on earth are obtained. For many purposes of analysis (erosion, building patterns, deforestation, glacier movement, etc.), these images of the same location are compared over time.
    [Show full text]
  • A New Data Model, Programming Interface, and Format Using HDF5
    NetCDF-4: A New Data Model, Programming Interface, and Format Using HDF5 Russ Rew, Ed Hartnett, John Caron UCAR Unidata Program Center Mike Folk, Robert McGrath, Quincey Kozial NCSA and The HDF Group, Inc. Final Project Review, August 9, 2005 THG, Inc. 1 Motivation: Why is this area of work important? While the commercial world has standardized on the relational data model and SQL, no single standard or tool has critical mass in the scientific community. There are many parallel and competing efforts to build these tool suites – at least one per discipline. Data interchange outside each group is problematic. In the next decade, as data interchange among scientific disciplines becomes increasingly important, a common HDF-like format and package for all the sciences will likely emerge. Jim Gray, Distinguished “Scientific Data Management in the Coming Decade,” Jim Gray, David Engineer at T. Liu, Maria A. Nieto-Santisteban, Alexander S. Szalay, Gerd Heber, Microsoft, David DeWitt, Cyberinfrastructure Technology Watch Quarterly, 1998 Turing Award Volume 1, Number 2, February 2005 winner 2 Preservation of scientific data … the ephemeral nature of both data formats and storage media threatens our very ability to maintain scientific, legal, and cultural continuity, not on the scale of centuries, but considering the unrelenting pace of technological change, from one decade to the next. … And that's true not just for the obvious items like images, documents, and audio files, but also for scientific images, … and MacKenzie Smith, simulations. In the scientific research community, Associate Director standards are emerging here and there—HDF for Technology at (Hierarchical Data Format), NetCDF (network the MIT Libraries, Common Data Form), FITS (Flexible Image Project director at Transport System)—but much work remains to be MIT for DSpace, a groundbreaking done to define a common cyberinfrastructure.
    [Show full text]
  • A Very Useful Enhancement for MSC Nastran and Patran
    Tips & Tricks HDF5: A Very Useful Enhancement for MSC Nastran and Patran et’s face it, as FEA engineers we’ve all been As a premier FEA solver that is widely used by the in that situation where the finite element Aerospace and Automotive industries, MSC Nastran L model has been prepared, boundary now takes advantage of the HDF5 file to manage conditions have been checked and the model has and store your FEA data. This can simplify your post been solved, but then it comes time to investigate processing tasks and make data management much the results. It’s not unusual on a large project to easier and simpler (see below). produce different outputs for different purposes: an XDB or OP2 for Patran, a Punch file to pass to So what is HDF5? It is an open source file format, Excel, and the f06 to read the printed output. In data model, and library developed by the HDF Group. addition, your company may have programs that read one of these files to perform special purpose HDF5 has three significant advantages compared to calculations. Managing all these files becomes a previous result file formats: 1) The HDF5 file is project of its own. smaller than XDB and OP2, 2) accessing results is significantly faster with HDF5, and 3) the input and Well let me tell you, there is a better way to do this!! output datablocks are stored in a single, high 82 | Engineering Reality Magazine Before HDF5 After HDF5 precision file. With input and output in the same file, party extensions are available for Python, .Net, and you eliminate confusion and potential errors keeping many other languages.
    [Show full text]
  • Importing and Exporting Data
    Chapter II-9 II-9Importing and Exporting Data Importing Data................................................................................................................................................ 117 Load Waves Submenu ............................................................................................................................ 119 Line Terminators...................................................................................................................................... 120 LoadWave Text Encodings..................................................................................................................... 120 Loading Delimited Text Files ........................................................................................................................ 120 Determining Column Formats............................................................................................................... 120 Date/Time Formats .................................................................................................................................. 121 Custom Date Formats ...................................................................................................................... 122 Column Labels ......................................................................................................................................... 122 Examples of Delimited Text ................................................................................................................... 123 The Load
    [Show full text]
  • Achieving High Performance I/O with HDF5
    Achieving High Performance I/O with HDF5 HDF5 Tutorial @ ECP Annual Meeting 2020 M. Scot Breitenfeld, Elena Pourmal The HDF Group Suren Byna, Quincey Koziol Lawrence Berkeley National Laboratory Houston, TX Feb 6th, 2020 exascaleproject.org Outline, Announcements and Resources INTRODUCTION 2 Tutorial Outline • Foundations of HDF5 • Parallel I/O with HDF5 • ECP HDF5 features and applications • Logistics – Live Google doc for instant questions and comments: https://tinyurl.com/uoxkwaq 3 https://tinyurl.com/uoxkwaq Announcements • HDF5 User Group meeting in June 2020 • The HDF Group Webinars https://www.hdfgroup.org/category/webinar/ – Introduction to HDF5 – HDF5 Advanced Features – HDF5 VOL connectors 4 https://tinyurl.com/uoxkwaq Resources • HDF5 home page: http://hdfgroup.org/HDF5/ • Latest releases: ØHDF5 1.10.6 https://portal.hdfgroup.org/display/support/Downloads ØHDF5 1.12.0 https://gamma.hdfgroup.org/ftp/pub/outgoing/hdf5_1.12/ • HDF5 repo: https://bitbucket.hdfgroup.org/projects/HDFFV/repos/hdf5/ • HDF5 Jira https://jira.hdfgroup.org • Documentation https://portal.hdfgroup.org/display/HDF5/HDF5 5 https://tinyurl.com/uoxkwaq FOUNDATIONS OF HDF5 Elena Pourmal 6 Foundations of HDF5 • Introduction to – HDF5 data model, software and architecture – HDF5 programming model • Overview of general best practices 7 https://tinyurl.com/uoxkwaq Why HDF5? • Have you ever asked yourself: – How do I organize and share my data? – How can I use visualization and other tools with my data? – What will happen to my data if I need to move my application
    [Show full text]
  • Efficient Lidar Point Cloud Data Encoding for Scalable Data Management Within the Hadoop Eco-System
    Efficient LiDAR point cloud data encoding for scalable data management within the Hadoop eco-system Vo, A. V., Hewage, C. N. L., Russo, G., Chauhan, N., Laefer, D. F., Bertolotto, M., Le-Khac, N-A., & Ofterdinger, U. (2020). Efficient LiDAR point cloud data encoding for scalable data management within the Hadoop eco- system. In IEEE BigData 2019 Los Angeles, CA, USA (pp. 5644-5653). IEEE . https://doi.org/10.1109/BigData47090.2019.9006044 Published in: IEEE BigData 2019 Los Angeles, CA, USA Document Version: Peer reviewed version Queen's University Belfast - Research Portal: Link to publication record in Queen's University Belfast Research Portal Publisher rights Copyright 2019 IEEE. This work is made available online in accordance with the publisher’s policies. Please refer to any applicable terms of use of the publisher. General rights Copyright for the publications made accessible via the Queen's University Belfast Research Portal is retained by the author(s) and / or other copyright owners and it is a condition of accessing these publications that users recognise and abide by the legal requirements associated with these rights. Take down policy The Research Portal is Queen's institutional repository that provides access to Queen's research output. Every effort has been made to ensure that content in the Research Portal does not infringe any person's rights, or applicable UK laws. If you discover content in the Research Portal that you believe breaches copyright or violates any law, please contact [email protected]. Download date:07. Oct. 2021 Efficient LiDAR point cloud data encoding for scalable data management within the Hadoop eco-system Anh Vu Vo Chamin Nalinda Lokugam Hewage Gianmarco Russo School of Computer Science School of Computer Science Department of Computer Science University College Dublin University College Dublin University of Salerno Dublin, Ireland Dublin, Ireland Fisciano, Italy [email protected] [email protected] [email protected] Neel Chauhan Debra F.
    [Show full text]
  • ECE 3574: Applied Software Design
    ECE 3574: Applied Software Design Message Serialization Today we are going to see various techniques for serializing objects, converting them to/from byte streams. The ability to exchange objects over files, pipes, sockets, and shared memory is a common task in concurrent programming. I Serialization (Marshalling or Pickling) I Deserialization (Unmarshalling or UnPickling) I QDataStream I Google Protocol buffers In general how do we make an object in memory persistent? We write it to a file. A file is a linear stream of bytes that may be I text-based: human-readable/editable, very portable and future proof, but we have to write parsers, they can be inefficient, can loose precision, take care to handle utf-8 I binary: machine-readable only, have to take care to make portable (Endianess), can easily get lost over time, but can preserve precision To recover the object we read the file back in. Example: ASCII text-based file Suppose we want to store some information about a task in some Todo app. struct Task { std::string description; bool completed; char priority; std::string context; }; Let’s store this as an ASCII text file. See example code: task_text_delimited.cpp There are a few standard text file formats I delimited files: common delimiters ‘,’, ‘:’, ‘|’ I INI files (originally from windows) I XML (heavy, usually overkill) I JSON (similar to xml but lighter) JSON can handle UTF encoding, arrays, many data types, and there are light-weight open-source parsing libraries for it (I like jsoncpp). Example: binary formatted file Binary files use sequences of bytes to store the same representation as in memory.
    [Show full text]
  • HDF5 Portable Parallel - Juelich.De I/O and I/O Data Formats Member of the Helmholtz-Association March March 14Th, 2017 Outline
    Parallel I/O and Portable Data Formats HDF5 Sebastian Lührs [email protected] Association - Jülich Supercomputing Centre Forschungszentrum Jülich GmbH Jülich, March 14th, 2017 Member Member of the Helmholtz Outline . Introduction . Structure of the HDF5 library . Terms and definitions . HDF5 - programming model and API . Creating/opening HDF5 files . Closing HDF5 files and other objects . HDF5 predefined datatypes . Creating dataspaces . Creating datasets . Writing/reading data . Row major / column major Association - . Partial I/O . Parallel HDF5 Member Member of the Helmholtz March 14th, 2017 2 What is HDF5? Hierarchical Data Format . API, data model and file format for I/O management . Tools suite for accessing data in HDF5 format Association - Member Member of the Helmholtz March 14th, 2017 3 HDF5 - Features . Supports parallel I/O . Self describing data model which allows the management of complex data sets . Portable file format . Available on a variety of platforms . Supports C, C++, Fortran 90 and Java . Pythonic interfaces also available . Provides tools to operate on HDF5 files and data Association - Member Member of the Helmholtz March 14th, 2017 4 Layers of the HDF5 Library Application Your Application Object API (C, F90, C++, Java, Python) Library Internals split HDF5 Library stdio MPI I/O Custom files Virtual File I/O Association - Storage device Member Member of the Helmholtz March 14th, 2017 5 File organization . HDF5 file structure corresponds in many respects to a Unix/Linux file system (fs) HDF5 Unix/Linux fs Group ↔ Directory Data set ↔ File Association - Member Member of the Helmholtz March 14th, 2017 6 Terminology File Container for storing data Group Structure which may contain HDF5 objects, e.g.
    [Show full text]