Introduction to HDF5
HDF & HDF-EOS Workshop XII
October 15, 2008

10/15/08

HDF & HDF-EOS Workshop XII 1

1
Topics Covered
- Introduce HDF5
- Describe HDF5 Data and Programming Models
- Walk Through Example Code

10/15/08

HDF & H...
For More Information …

All workshop slides will be available from:
http://hdfeos.org/workshops/ws12/workshop_twelve.php

...
What is HDF5?

HDF = Hierarchical Data Format
• Data model, library and file format for managing
data
• Tools for accessin...
Brief History of HDF
1987 At NCSA (University of Illinois), a task force formed to create an
architecture-independent form...
Why HDF5?

In one sentence ...

10/15/08

HDF & HDF-EOS Workshop XII 6

6
Answering big questions …

Matter and the universe

Life and nature

August 24, 2001

August 24, 2002

Total Column Ozone ...
… involves big data …

10/15/08

HDF & HDF-EOS Workshop XII 8

8
… varied data …

LCI Tutorial

10/15/08

Thanks to Mark
HDF & HDF-EOS Workshop XII 9 Miller, LLNL
9
… and complex relationships …
SNP Score
Contig Summaries
Discrepancies

Contig Qualities

Coverage Depth

Trace
Reads

Ali...
… on big computers …

… and small computers …
10/15/08

HDF & HDF-EOS Workshop XII 11
11
How do we…

• Describe our data?
• Read it? Store it? Find it? Share it? Mine it?
• Move it into, out of, and between comp...
Solution: HDF5!
• Can store all kinds of data in a variety of ways
• Runs on most systems
• Lots of tools to access data
•...
Structure of HDF5 Library

Applications
Object API (C, F90, C++, Java)
Library internals
Virtual file I/O
File or other “s...
HDF Tools
- HDFView and Java Products
- Command-line utilities (h5dump, h5ls, h5cc,
h5diff, h5repack)

10/15/08

HDF & HDF...
HDF5 Applications & Domains
Examples: Thermonuclear simulations
Product modeling
Data mining tools
Visualization tools
Cli...
Lots of Layers in HDF5!
“Ogres are like onions.”

Shrek  HDF5 Monster??
Just like Shrek, once you get to
know HDF5 you w...
The HDF5 Format

10/15/08

HDF & HDF-EOS Workshop XII 18
18
An HDF5 file is a container…

…into
which you
can put
your data
objects.

10/15/08

lat | lon | temp
----|-----|----12 | 2...
HDF5 Structures for Organizing Objects

“/” (root)
“foo”

3-D array

lat | lon | temp
----|-----|----12 | 23 | 3.1
15 | 24...
HDF5 Data Model
Primary Objects
• Groups
• Datasets

Additional ways to organize and annotate data
• Attributes
• Storage ...
HDF5 Dataset

Metadata

Data

Dataspace

Rank Dimensions
3

Dim_1 = 4
Dim_2 = 5
Dim_3 = 7

Datatype
Integer

Storage Info
...
Dataspaces
Two roles:
• Dataspace contains spatial info about a dataset
stored in a file
• Rank and dimensions
• Permanent...
Write – from memory to disk

memory

10/15/08

disk

HDF & HDF-EOS Workshop XII 24
24
Partial I/O
Move just part of a dataset
memory

disk

(a) Slab from a 2D array to the
corner of a smaller 2D array

Elemen...
Datatypes (array elements)
• Datatype – how to interpret a data element
• Permanent part of the dataset definition
• Two c...
Datatypes
• HDF5 atomic types include:
integer & float
user-definable (e.g., 13-bit integer)
variable length types (e.g...
HDF5 dataset: array of records
3

5

Dimensionality: 5 x 3
int8

int4

int16 2x3x2 array of float32

Datatype:

Record

10...
Properties

• Properties are characteristics of HDF5 objects
that can be modified
• Default properties handle most needs
•...
Special Storage Properties
Better subsetting
access time;
extensible

chunked

Improves storage
efficiency,
transmission s...
Attributes (optional)
• Attribute – data of the form “name = value”,
attached to an object
• Operations similar to dataset...
HDF5 Dataset (again)

Metadata

Data

Dataspace

Rank Dimensions
3

Dim_1 = 4
Dim_2 = 5
Dim_3 = 7

Datatype
Integer

Stora...
Groups
• A mechanism for organizing collections
• Every file starts with a root group
• Similar to UNIX directories
A
• Ca...
Path to HDF5 Object in a File

/ (root)
/x
/foo
/foo/temp
/foo/bar/temp

10/15/08

foo
temp

“/”

x

bar
temp

HDF & HDF-E...
Shared Objects
“/”

A

P

C

B
R

P

/A/P
/B/R
/C/P
10/15/08

HDF & HDF-EOS Workshop XII 35
35
Questions So Far?

10/15/08

HDF & HDF-EOS Workshop XII 36
Useful Tools For New Users
h5dump:
Tool to “dump” or display contents of HDF5 files

h5cc, h5c++, h5fc:
Scripts to compile...
H5dump Command-line Utility To View HDF5 File

h5dump [--header] [-a ] [-d <names>] [-g <names>]
[-l <names>] [-t <names>]...
Example of h5dump Output

HDF5 "dset.h5" {
GROUP "/" {
DATASET "dset" {
DATATYPE { H5T_STD_I32BE }
DATASPACE { SIMPLE ( 4,...
HDF5 Compile Scripts
• h5cc – HDF5 C compiler command
• h5fc – HDF5 F90 compiler command

• h5c++ – HDF5 C++ compiler comm...
Compile option: -show
-show: displays the compiler commands and options
without executing them
% h5cc –show Sample_c.c
gcc...
Browsing HDF5 Files with HDFView

10/15/08

HDF & HDF-EOS Workshop XII 42
HDFView

Structure of File

10/15/08

Contents
of Dataset

HDF & HDF-EOS Workshop XII 43
HDFView File Menu

10/15/08

HDF & HDF-EOS Workshop XII 44
10/15/08

HDF & HDF-EOS Workshop XII 45
Simple HDF5 File in HDFView
Right-click and select
“Open” with mouse
Right-click and select
“Show Properties”
with mouse

...
Simple HDF5 File in HDFView

10/15/08

HDF & HDF-EOS Workshop XII 47
HDF-EOS5 File in HDFView

10/15/08

HDF & HDF-EOS Workshop XII 48
Right-click and select
“Open As” with mouse

10/15/08

HDF & HDF-EOS Workshop XII 49
What you can’t see
with slides:
-Picture displayed instantly
-File size is 906,229,176

10/15/08

HDF & HDF-EOS Workshop X...
Introduction to
HDF5 Programming Model
and APIs

10/15/08

HDF & HDF-EOS Workshop XII 51
51
Operations Supported by the API
• Create objects (groups, datasets, attributes, complex data
types, …)
• Assign storage an...
General Programming Paradigm
• Properties of object are optionally defined
 Creation properties
 Access property lists

...
Order of Operations
• An order is imposed on operations by argument
dependencies
For Example:
A file must be opened before...
The General HDF5 API
• Currently C, Fortran 90, Java, and C++ bindings.
• C routines begin with prefix H5?
? is a characte...
HDF5 Defined Types
For portability, the HDF5 library has its own defined
types:
hid_t:
hsize_t:
hssize_t:

object identifi...
The HDF5 API
• For flexibility, the API is extensive
 300+ functions

Victronix
Swiss Army
Cybertool 34

• This can be da...
Basic Functions
H5Fcreate (H5Fopen)
H5Screate_simple
H5Dcreate (H5Dopen)
H5Dread, H5Dwrite
H5Dclose
H5Sclose
H5Fclose
10/1...
Other Common Functions
DataSpaces:

H5Sselect_hyperslab (Partial I/O)
H5Sselect_elements (Partial I/O)

Groups:

H5Gcreate...
High Level APIs

• Included along with the HDF5 library
• Simplify steps for creating, writing, and reading
objects
• Do n...
Example HDF5 Code

10/15/08

HDF & HDF-EOS Workshop XII 61
Steps to Create a File
1. Decide on special properties the file should have
•
•
•

Creation properties, like size of user ...
Code: Create a File
hid_t
herr_t

file_id;
status;

file_id = H5Fcreate ("file.h5", H5F_ACC_TRUNC,
H5P_DEFAULT, H5P_DEFAUL...
Dataset Components

Metadata

Data

Dataspace

Rank Dimensions
3

Dim_1 = 4
Dim_2 = 5
Dim_3 = 7

Datatype
Integer

Storage...
Steps to Create a Dataset
1. Define dataset characteristics
•
•
•

Dataspace - 4x6
Datatype – integer
Properties if needed...
HDF5 Pre-defined Datatype Identifiers
HDF5 defines* set of Datatype Identifiers per HDF5
session.
For example:
C Type

HDF...
Pre-defined File Datatype Identifiers
Examples:
H5T_IEEE_F64LE Eight-byte, little-endian, IEEE floating-point
H5T_STD_I32L...
Pre-defined Native Datatypes
Examples of predefined native types in C:
H5T_NATIVE_INT
H5T_NATIVE_FLOAT
H5T_NATIVE_UINT
H5T...
Dataset Creation Property List
Dataset creation property list: information on how to
organize data in storage.
Chunked

Ch...
Code: Create a Dataset
1
2
3

hid_t
hsize_t
herr_t

file_id, dataset_id, dataspace_id;
dims[2];
status;

4

file_id = H5Fc...
Example Code - H5Dwrite
Dataset Identifier from
H5Dcreate or H5Dopen

Memory Datatype

status = H5Dwrite (dataset_id, H5T_...
Example Code – H5Dwrite

status = H5Dwrite (dataset_id, H5T_NATIVE_INT, H5S_ALL, H5S_ALL,
H5P_DEFAULT, dset_data);

Data T...
Partial I/O
Memory Dataspace

H5S_ALL

File Dataspace (disk)

H5S_ALL

Get a Dataspace:
H5Screate_simple
H5Dget_space
Modi...
Example Code – H5Dread

status = H5Dread (dataset_id, H5T_NATIVE_INT,
H5S_ALL, H5S_ALL, H5P_DEFAULT, dset_rdata);

10/15/0...
High Level APIs: HDF5 Lite (H5LT)
#include "H5LT.h"
…
file_id = H5Fcreate (“file.h5", H5F_ACC_TRUNC,
H5P_DEFAULT, H5P_DEFA...
High Level APIs
•
•
•
•
•

10/15/08

HDF5 Lite
HDF5 Image
HDF5 Table
HDF5 Dimension Scales
HDF5 Packet Table

HDF & HDF-EO...
Example: Create a Group
“/” (root)
A

B

4x6 array of
integers

file.h5

10/15/08

HDF & HDF-EOS Workshop XII 77
77
Steps to Create a Group
1. Decide where to put it – “root group”
•

Obtain location ID

2. Decide name – “B”
3. Create gro...
Code: Create a Group
hid_t file_id, group_id;
...
/* Open “file.h5” */
file_id = H5Fopen (“file.h5”, H5F_ACC_RDWR,
H5P_DEF...
Thank you!
This work was supported by the Cooperative Agreement with the
National Aeronautics and Space Administration (NA...
Upcoming SlideShare
Loading in …5
×

Introduction to HDF5

404 views
239 views

Published on

This tutorial is designed for new HDF5 users. We will go over a brief history of HDF and HDF5 software, and will cover basic HDF5 Data Model objects and their properties; we will give an overview of the HDF5 Libraries and APIs, and discuss the HDF5 programming model. Simple C and Fortran examples, and Java tool HDFView will be used to illustrate HDF5 concepts.

Published in: Technology
0 Comments
0 Likes
Statistics
Notes
  • Be the first to comment

  • Be the first to like this

No Downloads
Views
Total views
404
On SlideShare
0
From Embeds
0
Number of Embeds
4
Actions
Shares
0
Downloads
11
Comments
0
Likes
0
Embeds 0
No embeds

No notes for slide
  • The CFD General Notation System (CGNS) provides a general, portable, and extensible standard for the storage and retrieval of computational fluid dynamics (CFD) analysis data. It consists of a collection of conventions, and free and open software implementing those conventions. It is self-descriptive, machine-independent, well-documented, and administered by an international steering committee.
    The CGNS implementation of SIDS, so-called MLL, was originally built using a file format called ADF (Advanced Data Format).  This format was based on a common file format system previously in use at McDonnell Douglas.  The ADF has worked extremely well, requiring little repair, upgrade, or maintenance over the last decade. However, ADF does not have parallel I/O or data compression capabilities, and does not have the support and tools that the storage format HDF5 offers.  HDF5, supported by The HDF Group, has rapidly grown to become a world-wide format standard for storing scientific data.  HDF5 has parallel capability as well as a broader support base than ADF. 
  • This shows that you can mix objects of different types according to your needs. Typically, there will be metadata stored with objects to indicate what type of object they are.
    Like HDF4, HDF5 has a grouping structure. The main difference is that every HDF5 file starts with a root group, whereas HDF4 doesn’t need any groups at all.
  • Data Array is an ordered collection of identically typed data items distinguished by their indices
    Metadata:
    Dataspace – Rank, dimensions; spatial info about dataset
    Datatype – Information on how to interpret your data
    Storage Properties – How array is organized
    Attributes – User-defined metadata (optional)
  • Here is an example of a basic HDF5 object.
    Notice that each element in the 3D array is a record with four values in it.
  • Data Array is an ordered collection of identically typed data items distinguished by their indices
    Metadata:
    Dataspace – Rank, dimensions; spatial info about dataset
    Datatype – Information on how to interpret your data
    Storage Properties – How array is organized
    Attributes – User-defined metadata (optional)
  • Like HDF4, HDF5 has a grouping structure.
    The main difference is that every HDF5 file starts with a root group, whereas HDF4 doesn’t need any groups at all.
  • To create this file, we would start by creating the file itself. When you create a file, the root group gets created with it. So every file has at least that one group.
  • Introduction to HDF5

    1. 1. Introduction to HDF5 HDF & HDF-EOS Workshop XII October 15, 2008 10/15/08 HDF & HDF-EOS Workshop XII 1 1
    2. 2. Topics Covered - Introduce HDF5 - Describe HDF5 Data and Programming Models - Walk Through Example Code 10/15/08 HDF & HDF-EOS Workshop XII 2 2
    3. 3. For More Information … All workshop slides will be available from: http://hdfeos.org/workshops/ws12/workshop_twelve.php 10/15/08 HDF & HDF-EOS Workshop XII 3
    4. 4. What is HDF5? HDF = Hierarchical Data Format • Data model, library and file format for managing data • Tools for accessing data in the HDF5 format 10/15/08 HDF & HDF-EOS Workshop XII 4
    5. 5. Brief History of HDF 1987 At NCSA (University of Illinois), a task force formed to create an architecture-independent format and library: AEHOO (All Encompassing Hierarchical Object Oriented format) Became HDF Early NASA adopted HDF for Earth Observing System project 1990’s 1996 DOE’s ASC (Advanced Simulation and Computing) Project began collaborating with the HDF group (NCSA) to create “Big HDF” (Increase in computing power of DOE systems at LLNL, LANL and Sandia National labs, required bigger, more complex data files). “Big HDF” became HDF5. 1998 HDF5 was released with support from National Labs, NASA, NCSA 2006 The HDF Group spun off from University of Illinois as non-profit corporation 10/15/08 HDF & HDF-EOS Workshop XII 5
    6. 6. Why HDF5? In one sentence ... 10/15/08 HDF & HDF-EOS Workshop XII 6 6
    7. 7. Answering big questions … Matter and the universe Life and nature August 24, 2001 August 24, 2002 Total Column Ozone (Dobson) 60 385 610 Weather and climate 10/15/08 HDF & HDF-EOS Workshop XII 7 7
    8. 8. … involves big data … 10/15/08 HDF & HDF-EOS Workshop XII 8 8
    9. 9. … varied data … LCI Tutorial 10/15/08 Thanks to Mark HDF & HDF-EOS Workshop XII 9 Miller, LLNL 9
    10. 10. … and complex relationships … SNP Score Contig Summaries Discrepancies Contig Qualities Coverage Depth Trace Reads Aligned bases Read quality Contig Percent match 10/15/08 HDF & HDF-EOS Workshop XII 10 10
    11. 11. … on big computers … … and small computers … 10/15/08 HDF & HDF-EOS Workshop XII 11 11
    12. 12. How do we… • Describe our data? • Read it? Store it? Find it? Share it? Mine it? • Move it into, out of, and between computers and repositories? • Achieve storage and I/O efficiency? • Give applications and tools easy access our data? 10/15/08 HDF & HDF-EOS Workshop XII 12 12
    13. 13. Solution: HDF5! • Can store all kinds of data in a variety of ways • Runs on most systems • Lots of tools to access data • Emphasis on standards (HDF-EOS, CGNS) • Library and format emphasis on I/O efficiency and storage 10/15/08 HDF & HDF-EOS Workshop XII 13
    14. 14. Structure of HDF5 Library Applications Object API (C, F90, C++, Java) Library internals Virtual file I/O File or other “storage” 10/15/08 HDF & HDF-EOS Workshop XII 14
    15. 15. HDF Tools - HDFView and Java Products - Command-line utilities (h5dump, h5ls, h5cc, h5diff, h5repack) 10/15/08 HDF & HDF-EOS Workshop XII 15 15
    16. 16. HDF5 Applications & Domains Examples: Thermonuclear simulations Product modeling Data mining tools Visualization tools Climate models Simulation, visualization, remote sensing… HDF-EOS Virtual File Layer (I/O Drivers) Stdio CGNS HDF5 Data Model & API Split Files MPI I/O Storage HDF5 format 10/15/08 File ASC Custom ? Split metadata File on parallel and raw data files file system User-defined device HDF & HDF-EOS Workshop XII 16 Communities
    17. 17. Lots of Layers in HDF5! “Ogres are like onions.” Shrek  HDF5 Monster?? Just like Shrek, once you get to know HDF5 you will really like it!! 10/15/08 HDF & HDF-EOS Workshop XII 17
    18. 18. The HDF5 Format 10/15/08 HDF & HDF-EOS Workshop XII 18 18
    19. 19. An HDF5 file is a container… …into which you can put your data objects. 10/15/08 lat | lon | temp ----|-----|----12 | 23 | 3.1 15 | 24 | 4.2 17 | 21 | 3.6 te let pa HDF & HDF-EOS Workshop XII 19 19
    20. 20. HDF5 Structures for Organizing Objects “/” (root) “foo” 3-D array lat | lon | temp ----|-----|----12 | 23 | 3.1 15 | 24 | 4.2 17 | 21 | 3.6 palette Table Raster image Raster image 10/15/08 2-D array HDF & HDF-EOS Workshop XII 20 20
    21. 21. HDF5 Data Model Primary Objects • Groups • Datasets Additional ways to organize and annotate data • Attributes • Storage and access properties Everything else is built from these parts. 10/15/08 HDF & HDF-EOS Workshop XII 21 21
    22. 22. HDF5 Dataset Metadata Data Dataspace Rank Dimensions 3 Dim_1 = 4 Dim_2 = 5 Dim_3 = 7 Datatype Integer Storage Info Attributes Time = 32.4 Chunked Pressure = 987 Compressed Temp = 56 10/15/08 HDF & HDF-EOS Workshop XII 22 22
    23. 23. Dataspaces Two roles: • Dataspace contains spatial info about a dataset stored in a file • Rank and dimensions • Permanent part of dataset definition Rank = 2 Dimensions = 4x6 • Partial I/0: Dataspace describes application’s data buffer and data elements participating in I/O Rank = 1 Dimension = 10 10/15/08 HDF & HDF-EOS Workshop XII 23 23
    24. 24. Write – from memory to disk memory 10/15/08 disk HDF & HDF-EOS Workshop XII 24 24
    25. 25. Partial I/O Move just part of a dataset memory disk (a) Slab from a 2D array to the corner of a smaller 2D array Elements in each must be same. 10/15/08 disk memory (b) Regular series of blocks from a 2D array to a contiguous sequence at a certain offset in a 1D array HDF & HDF-EOS Workshop XII 25 25
    26. 26. Datatypes (array elements) • Datatype – how to interpret a data element • Permanent part of the dataset definition • Two classes: atomic and compound 10/15/08 HDF & HDF-EOS Workshop XII 26 26
    27. 27. Datatypes • HDF5 atomic types include: integer & float user-definable (e.g., 13-bit integer) variable length types (e.g., strings) references to objects/dataset regions enumeration - names mapped to integers • HDF5 compound types Comparable to C structs (“records”) Members can be atomic or compound types 10/15/08 HDF & HDF-EOS Workshop XII 27 27
    28. 28. HDF5 dataset: array of records 3 5 Dimensionality: 5 x 3 int8 int4 int16 2x3x2 array of float32 Datatype: Record 10/15/08 HDF & HDF-EOS Workshop XII 28 28
    29. 29. Properties • Properties are characteristics of HDF5 objects that can be modified • Default properties handle most needs • By changing properties can take advantage of the more powerful features in HDF5 10/15/08 HDF & HDF-EOS Workshop XII 29
    30. 30. Special Storage Properties Better subsetting access time; extensible chunked Improves storage efficiency, transmission speed compressed Arrays can be extended in any direction extensible File B split file Dataset “Fred” File A Metadata for Fred 10/15/08 Metadata in one file, raw data in another Data for Fred HDF & HDF-EOS Workshop XII 30 30
    31. 31. Attributes (optional) • Attribute – data of the form “name = value”, attached to an object • Operations similar to dataset operations, but … Not extensible No compression or partial I/O • Can be overwritten, deleted, added during the “life” of a dataset 10/15/08 HDF & HDF-EOS Workshop XII 31 31
    32. 32. HDF5 Dataset (again) Metadata Data Dataspace Rank Dimensions 3 Dim_1 = 4 Dim_2 = 5 Dim_3 = 7 Datatype Integer Storage info Attributes Time = 32.4 Chunked Pressure = 987 Compressed Temp = 56 10/15/08 HDF & HDF-EOS Workshop XII 32 32
    33. 33. Groups • A mechanism for organizing collections • Every file starts with a root group • Similar to UNIX directories A • Can have attributes k 10/15/08 “/” B l m HDF & HDF-EOS Workshop XII 33 33 C
    34. 34. Path to HDF5 Object in a File / (root) /x /foo /foo/temp /foo/bar/temp 10/15/08 foo temp “/” x bar temp HDF & HDF-EOS Workshop XII 34 34
    35. 35. Shared Objects “/” A P C B R P /A/P /B/R /C/P 10/15/08 HDF & HDF-EOS Workshop XII 35 35
    36. 36. Questions So Far? 10/15/08 HDF & HDF-EOS Workshop XII 36
    37. 37. Useful Tools For New Users h5dump: Tool to “dump” or display contents of HDF5 files h5cc, h5c++, h5fc: Scripts to compile applications HDFView: Java browser to view HDF4 and HDF5 files 10/15/08 HDF & HDF-EOS Workshop XII 37
    38. 38. H5dump Command-line Utility To View HDF5 File h5dump [--header] [-a ] [-d <names>] [-g <names>] [-l <names>] [-t <names>] [-p] <file> --header Display header only; no data is displayed. -a <names> Display the specified attribute(s). -d <names> Display the specified dataset(s). -g <names> Display the specified group(s) and all the members. -l <names> Displays the value(s) of the specified soft link(s). -t <names> Display the specified named datatype(s). -p Display properties. <names> is one or more appropriate object names. 10/15/08 HDF & HDF-EOS Workshop XII 38
    39. 39. Example of h5dump Output HDF5 "dset.h5" { GROUP "/" { DATASET "dset" { DATATYPE { H5T_STD_I32BE } DATASPACE { SIMPLE ( 4, 6 ) / ( 4, 6 ) } DATA { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24 “/” } } } } 10/15/08 HDF & HDF-EOS Workshop XII 39 ‘dset’
    40. 40. HDF5 Compile Scripts • h5cc – HDF5 C compiler command • h5fc – HDF5 F90 compiler command • h5c++ – HDF5 C++ compiler command To compile: % h5cc h5prog.c % h5fc h5prog.f90 10/15/08 HDF & HDF-EOS Workshop XII 40 40
    41. 41. Compile option: -show -show: displays the compiler commands and options without executing them % h5cc –show Sample_c.c gcc -I/home/packages/hdf5_1.6.6/Linux_2.6/include -UH5_DEBUG_API -DNDEBUG -I/home/packages/szip/static/encoder/Linux2.6-gcc/include -D_LARGEFILE_SOURCE -D_LARGEFILE64_SOURCE -D_FILE_OFFSET_BITS=64 -D_POSIX_SOURCE -D_BSD_SOURCE -std=c99 -Wno-long-long -O -fomit-frame-pointer -finline-functions -c Sample_c.c gcc -std=c99 -Wno-long-long -O -fomit-frame-pointer -finline-functions -L/home/packages/szip/static/encoder/Linux2.6-gcc/lib Sample_c.o -L/home/packages/hdf5_1.6.6/Linux_2.6/lib /home/packages/hdf5_1.6.6/Linux_2.6/lib/libhdf5_hl.a /home/packages/hdf5_1.6.6/Linux_2.6/lib/libhdf5.a -lsz -lz -lm -Wl,-rpath -Wl,/home/packages/hdf5_1.6.6/Linux_2.6/lib 10/15/08 HDF & HDF-EOS Workshop XII 41 41
    42. 42. Browsing HDF5 Files with HDFView 10/15/08 HDF & HDF-EOS Workshop XII 42
    43. 43. HDFView Structure of File 10/15/08 Contents of Dataset HDF & HDF-EOS Workshop XII 43
    44. 44. HDFView File Menu 10/15/08 HDF & HDF-EOS Workshop XII 44
    45. 45. 10/15/08 HDF & HDF-EOS Workshop XII 45
    46. 46. Simple HDF5 File in HDFView Right-click and select “Open” with mouse Right-click and select “Show Properties” with mouse 10/15/08 HDF & HDF-EOS Workshop XII 46
    47. 47. Simple HDF5 File in HDFView 10/15/08 HDF & HDF-EOS Workshop XII 47
    48. 48. HDF-EOS5 File in HDFView 10/15/08 HDF & HDF-EOS Workshop XII 48
    49. 49. Right-click and select “Open As” with mouse 10/15/08 HDF & HDF-EOS Workshop XII 49
    50. 50. What you can’t see with slides: -Picture displayed instantly -File size is 906,229,176 10/15/08 HDF & HDF-EOS Workshop XII 50
    51. 51. Introduction to HDF5 Programming Model and APIs 10/15/08 HDF & HDF-EOS Workshop XII 51 51
    52. 52. Operations Supported by the API • Create objects (groups, datasets, attributes, complex data types, …) • Assign storage and I/O properties to objects • Perform complex subsetting during read/write • Use variety of I/O “devices” (parallel, remote, etc.) • Transform data during I/O • Make inquiries on file and object structure, content, properties 10/15/08 HDF & HDF-EOS Workshop XII 52 52
    53. 53. General Programming Paradigm • Properties of object are optionally defined  Creation properties  Access property lists • Object is opened or created • Object is accessed, possibly many times • Object is closed 10/15/08 HDF & HDF-EOS Workshop XII 53 53
    54. 54. Order of Operations • An order is imposed on operations by argument dependencies For Example: A file must be opened before a dataset -becausethe dataset open call requires a file handle as an argument. • Objects can be closed in any order. 10/15/08 HDF & HDF-EOS Workshop XII 54 54
    55. 55. The General HDF5 API • Currently C, Fortran 90, Java, and C++ bindings. • C routines begin with prefix H5? ? is a character corresponding to the type of object the function acts on Example Functions: H5D : Dataset interface H5F : File interface e.g., H5Dread e.g., H5Fopen H5S : dataSpace interface e.g., H5Sclose 10/15/08 HDF & HDF-EOS Workshop XII 55 55
    56. 56. HDF5 Defined Types For portability, the HDF5 library has its own defined types: hid_t: hsize_t: hssize_t: object identifiers (native integer) size used for dimensions (unsigned long or unsigned long long) for specifying coordinates and sometimes for dimensions (signed long or signed long long) herr_t: function return value hvl_t: variable length datatype For C, include hdf5.h in your HDF5 application. 10/15/08 HDF & HDF-EOS Workshop XII 56 56
    57. 57. The HDF5 API • For flexibility, the API is extensive  300+ functions Victronix Swiss Army Cybertool 34 • This can be daunting… but there is hope  A few functions can do a lot  Start simple  Build up knowledge as more features are needed 10/15/08 HDF & HDF-EOS Workshop XII 57 57
    58. 58. Basic Functions H5Fcreate (H5Fopen) H5Screate_simple H5Dcreate (H5Dopen) H5Dread, H5Dwrite H5Dclose H5Sclose H5Fclose 10/15/08 create (open) File create dataSpace create (open) Dataset access Dataset close Dataset close dataSpace close File HDF & HDF-EOS Workshop XII 58
    59. 59. Other Common Functions DataSpaces: H5Sselect_hyperslab (Partial I/O) H5Sselect_elements (Partial I/O) Groups: H5Gcreate, H5Gopen, H5Gclose Attributes: H5Acreate, H5Aopen_name, H5Aclose, H5Aread, H5Awrite Property lists: H5Pcreate, H5Pclose H5Pset_chunk, H5Pset_deflate 10/15/08 HDF & HDF-EOS Workshop XII 59
    60. 60. High Level APIs • Included along with the HDF5 library • Simplify steps for creating, writing, and reading objects • Do not entirely ‘wrap’ HDF5 library 10/15/08 HDF & HDF-EOS Workshop XII 60
    61. 61. Example HDF5 Code 10/15/08 HDF & HDF-EOS Workshop XII 61
    62. 62. Steps to Create a File 1. Decide on special properties the file should have • • • Creation properties, like size of user block Access properties, such as metadata cache size Use default properties (H5P_DEFAULT) 2. Create property lists, if necessary 3. Create the file 4. Close the file and the property lists, as needed 10/15/08 HDF & HDF-EOS Workshop XII 62 62
    63. 63. Code: Create a File hid_t herr_t file_id; status; file_id = H5Fcreate ("file.h5", H5F_ACC_TRUNC, H5P_DEFAULT, H5P_DEFAULT); status = H5Fclose (file_id); “/” (root) Note: Return codes not checked for errors in code samples. 10/15/08 HDF & HDF-EOS Workshop XII 63 63
    64. 64. Dataset Components Metadata Data Dataspace Rank Dimensions 3 Dim_1 = 4 Dim_2 = 5 Dim_3 = 7 Datatype Integer Storage info Attributes Time = 32.4 Chunked Pressure = 987 Compressed Temp = 56 10/15/08 HDF & HDF-EOS Workshop XII 64 64
    65. 65. Steps to Create a Dataset 1. Define dataset characteristics • • • Dataspace - 4x6 Datatype – integer Properties if needed, or use H5P_DEFAULT 2. Decide where to put it • Obtain location ID: - Group ID puts it in a Group - File ID puts it in Root Group “/” (root) A 3. Create dataset in file 4. Close everything 10/15/08 HDF & HDF-EOS Workshop XII 65 65
    66. 66. HDF5 Pre-defined Datatype Identifiers HDF5 defines* set of Datatype Identifiers per HDF5 session. For example: C Type HDF5 File Type HDF5 Memory Type int H5T_STD_I32BE H5T_STD_I32LE H5T_NATIVE_INT float H5T_IEEE_F32BE H5T_IEEE_F32LE H5T_NATIVE_FLOAT double H5T_IEEE_F64BE H5T_IEEE_F64LE H5T_NATIVE_DOUBLE * Value of datatype is NOT fixed 10/15/08 HDF & HDF-EOS Workshop XII 66
    67. 67. Pre-defined File Datatype Identifiers Examples: H5T_IEEE_F64LE Eight-byte, little-endian, IEEE floating-point H5T_STD_I32LE Four-byte, little-endian, signed two's complement integer Architecture* Programming Type NOTE: What you see in the file. Name is the same everywhere and explicitly defines a datatype. *STD= “An architecture with a semi-standard type like 2’s complement integer, unsigned integer…” 10/15/08 HDF & HDF-EOS Workshop XII 67
    68. 68. Pre-defined Native Datatypes Examples of predefined native types in C: H5T_NATIVE_INT H5T_NATIVE_FLOAT H5T_NATIVE_UINT H5T_NATIVE_LONG H5T_NATIVE_CHAR (int) (float ) (unsigned int) (long ) (char ) NOTE: Memory types. Different for each machine. Used for reading/writing. 10/15/08 HDF & HDF-EOS Workshop XII 68
    69. 69. Dataset Creation Property List Dataset creation property list: information on how to organize data in storage. Chunked Chunked & compressed H5P_DEFAULT: contiguous 10/15/08 HDF & HDF-EOS Workshop XII 69 69
    70. 70. Code: Create a Dataset 1 2 3 hid_t hsize_t herr_t file_id, dataset_id, dataspace_id; dims[2]; status; 4 file_id = H5Fcreate (”file.h5", H5F_ACC_TRUNC, H5P_DEFAULT, H5P_DEFAULT); Create a 5 dataspace= 4; dims[0] 6 7 rank dims[1] = 6; dataspace_id = H5Screate_simple (2, dims, NULL); Create a dataset 8 current dims pathname datatype dataset_id = H5Dcreate(file_id,”A",H5T_STD_I32BE, dataspace_id, H5P_DEFAULT); dataspace Terminate access to dataset, dataspace, file 9 status = H5Dclose (dataset_id); 10 status = H5Sclose (dataspace_id); 11 status = H5Fclose (file_id); 10/15/08 property list (default) HDF & HDF-EOS Workshop XII 70 70
    71. 71. Example Code - H5Dwrite Dataset Identifier from H5Dcreate or H5Dopen Memory Datatype status = H5Dwrite (dataset_id, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, dset_data); 10/15/08 HDF & HDF-EOS Workshop XII 71
    72. 72. Example Code – H5Dwrite status = H5Dwrite (dataset_id, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, dset_data); Data Transfer Property List (MPI I/O, Transformations, …) Memory Dataspace File Dataspace H5S_ALL selects entire dataspace 10/15/08 HDF & HDF-EOS Workshop XII 72
    73. 73. Partial I/O Memory Dataspace H5S_ALL File Dataspace (disk) H5S_ALL Get a Dataspace: H5Screate_simple H5Dget_space Modify Dataspace: H5Sselect_hyperslab H5Sselect_elements 10/15/08 HDF & HDF-EOS Workshop XII 73
    74. 74. Example Code – H5Dread status = H5Dread (dataset_id, H5T_NATIVE_INT, H5S_ALL, H5S_ALL, H5P_DEFAULT, dset_rdata); 10/15/08 HDF & HDF-EOS Workshop XII 74
    75. 75. High Level APIs: HDF5 Lite (H5LT) #include "H5LT.h" … file_id = H5Fcreate (“file.h5", H5F_ACC_TRUNC, H5P_DEFAULT, H5P_DEFAULT); status = H5LTmake_dataset (file_id,“A", 2, dims, H5T_STD_I32BE, data); status = H5Fclose (file_id); 10/15/08 HDF & HDF-EOS Workshop XII 75
    76. 76. High Level APIs • • • • • 10/15/08 HDF5 Lite HDF5 Image HDF5 Table HDF5 Dimension Scales HDF5 Packet Table HDF & HDF-EOS Workshop XII 76
    77. 77. Example: Create a Group “/” (root) A B 4x6 array of integers file.h5 10/15/08 HDF & HDF-EOS Workshop XII 77 77
    78. 78. Steps to Create a Group 1. Decide where to put it – “root group” • Obtain location ID 2. Decide name – “B” 3. Create group in file 4. (Eventually) close the group. 10/15/08 HDF & HDF-EOS Workshop XII 78 78
    79. 79. Code: Create a Group hid_t file_id, group_id; ... /* Open “file.h5” */ file_id = H5Fopen (“file.h5”, H5F_ACC_RDWR, H5P_DEFAULT); /* Create group "/B" in file. */ group_id = H5Gcreate (file_id,"B",0); Size hint for number of bytes to store names of objects. 0=default /* Close group and file. */ status = H5Gclose (group_id); status = H5Fclose (file_id); 10/15/08 HDF & HDF-EOS Workshop XII 79 79
    80. 80. Thank you! This work was supported by the Cooperative Agreement with the National Aeronautics and Space Administration (NASA) under NASA grant NNX06AC83A and NNX08A077A. Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of NASA. 10/15/08 HDF & HDF-EOS Workshop XII 80

    ×