Property (H5P)

Object property lists

Description

Use the MATLAB^® HDF5 property interface, H5P, to control and access information about object properties.

General Property List Operations

H5P.close

Close property list

H5P.close(plistID) terminates access to the property list specified by plistID.

H5P.copy

Copy of property list

newplist = H5P.copy(plistID) returns a copy of the property list specified by plistID.

H5P.create

Create new property list

plist = H5P.create(classID) creates a new property list as an instance of the property list class specified by classID. The classID argument can also be an instance of a property list class.

 Details

Specify classID as one of these string scalars or character vectors, or their corresponding constant values:

• "H5P_ATTRIBUTE_CREATE"

• "H5P_DATASET_ACCESS"

• "H5P_DATASET_CREATE"

• "H5P_DATASET_XFER"

• "H5P_DATATYPE_CREATE"

• "H5P_DATATYPE_ACCESS"

• "H5P_FILE_MOUNT"

• "H5P_FILE_CREATE"

• "H5P_FILE_ACCESS"

• "H5P_GROUP_CREATE"

• "H5P_GROUP_ACCESS"

• "H5P_LINK_CREATE"

• "H5P_LINK_ACCESS"

• "H5P_OBJECT_COPY"

• "H5P_OBJECT_CREATE"

• "H5P_STRING_CREATE"

H5P.get_class

Property list class

plistClass = H5P.get_class(plistID) returns the property list class for the property list specified by plistID.

Generic Property List Operations

H5P.close_class

Close property list class

H5P.close_class(classID) closes the property list class specified by classID.

H5P.equal

Determine equality of property lists

output = H5P.equal(plistID1,plistID2) returns a positive value if the two property lists specified are equal, and 0 if they are not. A negative return value indicates failure.

H5P.exist

Determine if specified property exists in property list

output = H5P.exist(propID,propname) returns a positive value if the property specified by the string scalar or character vector propname exists within the property list or class specified by propID, and 0 if it does not. A negative return value indicates failure.

H5P.get

Value of specified property in property list

value = H5P.get(plistID,propname) retrieves a copy of the value of the property specified by propname in the property list specified by plistID. Specify propname as a string scalar or character vector. The H5P.get function returns the property as an array of uint8 values. You might need to cast the value to an appropriate datatype to get a meaningful result.

It is recommended to use alternative functions like H5P.get_chunk, H5P.get_layout, H5P.get_size etc., where available, to get values for the common property names.

H5P.get_class_name

Name of property list class

classname = H5P.get_class_name(classID) retrieves the name of the generic property list class and returns it as a character vector classname. If no class is found, H5P.get_class_name returns an empty character vector.

H5P.get_class_parent

Identifier for parent class

pclassObj = H5P.get_class_parent(pclassID) returns an identifier to the parent class object of the property class specified by pclassID.

H5P.get_nprops

Query number of properties in property list or class

numProps = H5P.get_nprops(propID) returns the number of properties in the property list or class specified by propID.

H5P.get_size

Query size of property value in bytes

sz = H5P.get_size(propID,propname) returns the size in bytes of the property specified by propname in the property list or property class specified by propID. Specify propname as a string scalar or character vector.

H5P.isa_class

Determine if property list is member of class

output = H5P.isa_class(plistID,pclassID) returns a positive value if the property list specified by plistID is a member of the class specified by pclassID, and 0 if it is not. A negative return value indicates failure.

H5P.iterate

Iterate over properties in property list

[output,idxOut] = H5P.iterate(propID,idxIn,fnc) executes the specified function fnc on each property in the property object specified in propID.

 Details

Input Arguments

• propID — Identifier of the property list or property class.

• idxIn — Index at which iteration begins.

• fnc — Callback function handle with this signature: status = fnc(f_propID,propname).

  Input Arguments for fnc

  • f_propID — Populated with propID input argument.

  • propname — Name of current property.

  Output Arguments for fnc

  • status — Status indicator, interpreted as follows:

    • zero — Continues with the iteration, or populates output output argument if all members have been processed.

    • nonzero — Stops the iteration and populates the output output argument.

Output Arguments

• output — Populated by the final status output argument of fnc at the end of the iteration and interpreted as follows:

  • zero — Success. All properties were processed.

  • positive — Short-circuit success. The value is the index value of the last property that was processed.

  • negative — Failure.

• idxOut — Property index of the last property processed. This value can be used to resume an interrupted iteration.

H5P.set

Set property list value

H5P.set(plistID,propname,value) sets the value of the property specified by propname in the property list specified by plistID to the value specified in value.

It is recommended to use alternative functions like H5P.set_chunk, H5P.set_layout, H5P.set_size, etc., where available, to set values for the common property names.

 Details

Input Arguments

• plistID — Property list identifier.

• propname — Property name, specified as a string scalar or character vector.

• value — Value to set the property propname. The datatype of value must be uint8.

Dataset Access, Memory, and Transfer Properties

H5P.get_btree_ratios

B-tree split ratios

[left,middle,right] = H5P.get_btree_ratios(plistID) returns the B-tree split ratios for the dataset transfer property list specified by plistID. The left output specifies the B-tree split ratio for left-most nodes, right corresponds to the right-most nodes and lone nodes, and middle corresponds to all other nodes.

H5P.get_chunk_cache

Raw data chunk cache parameters

[nslots,nbytes,w0] = H5P.get_chunk_cache(daplID) returns the number of chunk slots in the raw data chunk cache hash table (nslots), the maximum possible number of bytes in the raw data chunk cache (nbytes), and the preemption policy value w0 of a dataset access property list specified by daplID.

H5P.get_edc_check

Determine if error detection is enabled

output = H5P.get_edc_check(plistID) queries the dataset transfer property list specified by plistID to determine whether error detection is enabled for data read operations. H5P.get_edc_check returns "H5Z_ENABLE_EDC" if error detection is enabled, and "H5Z_DISABLE_EDC" if it is not.

H5P.get_hyper_vector_size

Number of I/O vectors

sz = H5P.get_hyper_vector_size(dxplID) returns the number of I/O vectors to be read or written in hyperslab I/O.

H5P.set_btree_ratios

Set B-tree split ratios for dataset transfer

H5P.set_btree_ratios(plistID,left,middle,right) sets the B-tree split ratios for the dataset transfer property list specified by plistID. The left argument specifies the B-tree split ratio for left-most nodes, right specifies the B-tree split ratio for right-most nodes and lone nodes, and middle specifies the B-tree split ratio for all other nodes.

H5P.set_chunk_cache

Set raw data chunk cache parameters

H5P.set_chunk_cache(daplID,nslots,nbytes,w0) sets the number of elements nslots, the total number of bytes nbytes, and the preemption policy value w0 in the raw data chunk cache.

H5P.set_edc_check

Enable error detection for dataset transfer

H5P.set_edc_check(dtplID,check) sets the dataset transfer property list specified by dtplID to enable or disable error detection when reading data. Specify check as a string scalar or character vector containing "H5Z_ENABLE_EDC" or "H5Z_DISABLE_EDC".

H5P.set_hyper_vector_size

Set number of I/O vectors for hyperslab I/O

H5P.set_hyper_vector_size(dxplID,size) sets the number of I/O vectors to be accumulated in memory before being issued to the lower levels of the HDF5 library for reading or writing the actual data.

 Details

Input Arguments

• dxplID — Dataset transfer property list identifier.

• size — Number of I/O vectors to accumulate in memory for I/O operations.

Dataset Creation Properties

H5P.all_filters_avail

Determine availability of all filters

output = H5P.all_filters_avail(dcplID) returns a positive value if all of the filters set in the dataset creation property list dcplID are currently available, and 0 otherwise.

H5P.fill_value_defined

Determine if fill value is defined

fvstatus = H5P.fill_value_defined(dcplID) determines whether a fill value is defined in the dataset creation property list specified by dcplID.

 Details

The output, fvstatus, can have any of these values:

• "H5D_FILL_VALUE_UNDEFINED"

• "H5D_FILL_VALUE_DEFAULT"

• "H5D_FILL_VALUE_USER_DEFINED"

H5P.get_alloc_time

Timing of storage space allocation

allocTime = H5P.get_alloc_time(dcplID) retrieves the timing for storage space allocation from the dataset creation property list specified by dcplID.

 Details

The allocTime output can have any of these values:

• "H5D_ALLOC_TIME_DEFAULT"

• "H5D_ALLOC_TIME_EARLY"

• "H5D_ALLOC_TIME_INCR"

• "H5D_ALLOC_TIME_LATE"

H5P.get_chunk

Size of chunks

[rank,chunkdims] = H5P.get_chunk(dcplID) retrieves the chunk size of a dataset with dataset creation property list specified by dcplID.

Note

The HDF5 C library uses C-style ordering for multidimensional arrays, while MATLAB uses FORTRAN-style ordering. The chunkDims parameter assumes C-style ordering. For more information, see Report Data Set Dimensions.

H5P.get_external

Information about external file

[name,offset,size] = H5P.get_external(dcplID,idx) returns information about the external file specified by the dataset creation property list dcplID.

 Details

Input Arguments

• dcplID — Dataset creation property list identifier.

• idx — External file index, specified as a number from 0 to N – 1, where N is the value returned by H5P.get_external_count.

Output Arguments

• name — Name of the external file, limited to 2048 characters.

• offset — Location in bytes from the beginning of the external file where the data starts.

• size — Size of external data.

H5P.get_external_count

Count of external files

numfiles = H5P.get_external_count(dcplID) returns the number of external files for the dataset creation property list specified by dcplID.

H5P.get_fill_time

Time when fill values are written to dataset

fillTime = H5P.get_fill_time(dcplID) returns the time when fill values are written to the dataset creation property list specified by dcplID.

 Details

The fillTime output can be one of these values:

• "H5D_FILL_TIME_IFSET"

• "H5D_FILL_TIME_ALLOC"

• "H5D_FILL_TIME_NEVER"

H5P.get_fill_value

Dataset fill value

value = H5P.get_fill_value(dcplID,typeID) returns the dataset fill value defined in the dataset creation property list specified by dcplID. The typeID argument specifies the datatype of the returned fill value.

H5P.get_filter

Information about filter in pipeline

[filter,flags,values,name] = H5P.get_filter(dcplID,filterIdx) returns information about the filter, specified by its filter index, in the filter pipeline specified by the property list identifier dcplID. This syntax corresponds to the H5Pget_filter interface in version 1.6 of the HDF5 C library.

[filter,flags,values,name,config] = H5P.get_filter(dcplID,filterIdx) returns information about the filter, specified by its filter index, in the filter pipeline, specified by the property list with which it is associated. It also returns information about the filter. Consult the HDF5 documentation for H5Zget_filter_info for information about config. This syntax corresponds to the H5Pget_filter interface in version 1.8 of the HDF5 C library.

 Details

Input Arguments

• dcplID — Dataset or group creation property list identifier.

• filterIdx — Sequence number within the filter pipeline of the filter for which information is sought.

Output Arguments

• filter — Identifier of filter, interpreted as one of the following:

  • "H5Z_FILTER_DEFLATE" — Data compression filter, employing the gzip algorithm.

  • "H5Z_FILTER_SHUFFLE" — Data shuffling filter.

  • "H5Z_FILTER_FLETCHER32" — Error detection filter, employing the Fletcher32 checksum algorithm.

  • "H5Z_FILTER_SZIP" — Data compression filter, employing the SZIP algorithm.

  • "H5Z_FILTER_NBIT" — Data compression filter, employing the N-bit algorithm.

  • "H5Z_FILTER_SCALEOFFSET" — Data compression filter, employing the scale-offset algorithm.

• flags — Bit vector specifying certain general properties of the filter.

• values — Auxiliary data for the filter.

• name — Name of the filter.

• config — Bit field encoding the returned filter information.

H5P.get_filter_by_id

Information about specified filter

[flags,values,name,config] = H5P.get_filter_by_id(dcplID,filter) returns information about the filter specified by filter.

For custom third-party filters, specify filter as the numeric filter identifier assigned by The HDF Group.

 Details

Input Arguments

• dcplID — Dataset or group creation property list identifier.

• filter — Filter identifier.

Output Arguments

• flags — Bit vector specifying certain general properties of the filter.

• values — Auxiliary data for the filter.

• name — Name of the filter.

• config — Bit field encoding the returned filter information.

H5P.get_layout

Determine layout of raw data for dataset

layout = H5P.get_layout(dcplID) returns the layout of the raw data for the dataset specified by dataset creation property list identifier dcplID.

 Details

The H5P.get_layout function returns one of these values:

• "H5D_COMPACT" — Raw data is stored in the object header in the file.

• "H5D_CONTIGUOUS" — Raw data is stored separately from the object header in one contiguous chunk in the file.

• "H5D_CHUNKED" — Raw data is stored separately from the object header in chunks in separate locations in the file.

• "H5D_VIRTUAL" — Raw data is drawn from multiple datasets in different files.

H5P.get_nfilters

Number of filters in pipeline

numfilters = H5P.get_nfilters(dcplID) returns the number of filters defined in the filter pipeline associated with the dataset creation property list identifier dcplID.

H5P.modify_filter

Modify filter in pipeline

H5P.modify_filter(dcplID,filter,flags,cdValues) modifies the filter in the filter pipeline.

 Details

Input Arguments

• dcplID — Dataset creation property list identifier.

• filter — Identifier of filter to modify in filter pipeline, specified as one of these string scalars or character vectors or their numeric equivalents:

  • "H5Z_FILTER_DEFLATE"

  • "H5Z_FILTER_SHUFFLE"

  • "H5Z_FILTER_FLETCHER32"

  • "H5Z_FILTER_SZIP" (since R2024b)

  For custom third-party filters, specify filter as the numeric filter identifier assigned by The HDF Group.

• flags — Constant that specifies the behavior when the filter fails using these valid values:

  • "H5Z_FLAG_OPTIONAL" — Filter is excluded from the pipeline for the chunk where the filter failed and the filter does not participate in the pipeline during a subsequent read of the chunk.

  • "H5Z_FLAG_MANDATORY" — The HDF5 library issues an error upon filter failure. The library writes all chunks processed by the filter before the failure occurred.

• cdValues — Array containing auxiliary data for the filter.

H5P.remove_filter

Remove filter from property list

H5P.remove_filter(dcplID,filter) removes the specified filter from the filter pipeline. dcplID is the dataset creation property list identifier.

H5P.set_alloc_time

Set timing for storage space allocation

H5P.set_alloc_time(dcplID,allocTime) sets the timing for the allocation of storage space for the raw data of a dataset.

 Details

Input Arguments

• dcplID — Dataset creation property list identifier.

• allocTime — Timing for storage allocation, specified as one of these string scalars or character vectors:

  • "H5D_ALLOC_TIME_DEFAULT"

  • "H5D_ALLOC_TIME_EARLY"

  • "H5D_ALLOC_TIME_INC"

  • "H5D_ALLOC_TIME_LATE"

H5P.set_chunk

Set chunk size

H5P.set_chunk(dcplID,chunkDims) sets the size of the chunks used to store a chunked layout dataset.

 Details

Input Arguments

• dcplID — Dataset creation property list identifier.

• chunkDims — Array specifying the size, in dataset elements, of each chunk.

Note

The HDF5 C library uses C-style ordering for multidimensional arrays, while MATLAB uses FORTRAN-style ordering. The chunkDims parameter assumes C-style ordering. For more information, see Report Data Set Dimensions.

H5P.set_deflate

Set compression method and compression level

H5P.set_deflate(dcplID,level) sets the compression method for the dataset creation property list specified by dcplID to H5D_COMPRESS_DEFLATE. Specify the compression level level as an integer scalar value between 0 and 9. A value of 0 indicates no compression. A value of 1 indicates the least compression, and a value of 9 indicates the most.

H5P.set_external

Add additional file to external file list

H5P.set_external(dcplID,name,offset,nbytes) adds the external file specified by name to the list of external files in the dataset creation property list, dcplID.

 Details

Input Arguments

• dcplID — Dataset creation property list identifier.

• name — Name of the external file.

• offset — Location in bytes where the data starts relative to the beginning of the file.

• nbytes — Number of bytes reserved in the file for the data, and may also be given as "H5F_UNLIMITED", in which case the external file may be of unlimited size.

H5P.set_fill_time

Set time when fill values are written to dataset

H5P.set_fill_time(dcplID,fillTime) sets the timing for writing fill values to a dataset in the dataset creation property list specified by dcplID.

 Details

Specify fillTime as one of these values:

• "H5D_FILL_TIME_IFSET"

• "H5D_FILL_TIME_ALLOC"

• "H5D_FILL_TIME_NEVER"

H5P.set_fill_value

Set fill value for dataset creation property list

H5P.set_fill_value(dcplID,typeID,value) sets the fill value for the dataset creation property list specified by dcplID.

 Details

Input Arguments

• dcplID — Dataset creation property list identifier.

• typeID — Datatype of the fill value.

• value — Fill value. Setting value to an empty double array ([]) indicates that the fill value is undefined.

H5P.set_filter

Add filter to filter pipeline

H5P.set_filter(dcplID,filter,flags,cdValues) adds the specified filter and corresponding properties to the end of an output filter pipeline in plistID.

 Details

Input Arguments

• dcplID — Dataset creation property list identifier.

• filter — Identifier of filter to add to the end of the filter pipeline, specified as one of these string scalars or character vectors or their numeric equivalents:

  • "H5Z_FILTER_DEFLATE"

  • "H5Z_FILTER_SHUFFLE"

  • "H5Z_FILTER_FLETCHER32"

  • "H5Z_FILTER_SZIP" (since R2024b)

  For custom third-party filters, specify filter as the numeric filter identifier assigned by The HDF Group.

• flags — Constant that specifies the behavior when the filter fails. Valid values are:

  • "H5Z_FLAG_OPTIONAL" — Filter is excluded from the pipeline for the chunk in which the filter failed. The filter does not participate in the pipeline during a subsequent read of the chunk.

  • "H5Z_FLAG_MANDATORY" — The HDF5 library issues an error upon filter failure. The library writes all chunks processed by the filter before the failure occurred.

• cdValues — Array containing auxiliary data for the filter.

H5P.set_fletcher32

Set Fletcher32 checksum filter in dataset creation

H5P.set_fletcher32(dcplID) sets the 32-bit Fletcher checksum filter in the dataset creation property list specified by dcplID. The dataset creation property list must have chunking enabled.

H5P.set_layout

Set type of storage for dataset

H5P.set_layout(dcplID,layout) sets the type of storage used to store the raw data for the dataset creation property list specified by dcplID.

 Details

Specify layout as one of these values:

• "H5D_COMPACT" — Store raw data in the dataset object header in file. This should only be used for datasets with small amounts of raw data.

• "H5D_CONTIGUOUS" — Store aw data separately from the object header in one contiguous chunk in the file.

• "H5D_CHUNKED" — Store raw data separately from the object header in chunks in separate locations in the file.

• "H5D_VIRTUAL" — Draw raw data from multiple datasets in different files.

H5P.set_nbit

Set N-Bit filter

H5P.set_nbit(dcplID) sets the N-Bit filter "H5Z_FILTER_NBIT" for the dataset property list specified by dcplID.

H5P.set_scaleoffset

Set Scale-Offset filter

H5P.set_scaleoffset(dcplID,scaleType,scaleFactor) sets the scale offset filter "H5Z_FILTER_SCALEOFFSET" for a dataset specified by dcplID. Chunking must already be enabled on the dataset creation property list.

 Details

• dcplID — Dataset creation property list identifier.

• scaleType — Scale type, specified as one of these enumerated values:

  • "H5Z_SO_INT" — Enumerated value for integer datatypes.

  • "H5Z_SO_FLOAT_DSCALE" — Enumerated value for floating-point datatypes.

• scaleFactor — Scale factor.

H5P.set_shuffle

Set shuffle filter

H5P.set_shuffle(dcplID) sets the shuffle filter, H5Z_FILTER_SHUFFLE, in the dataset property list specified by dcplID. Compression must be enabled on the dataset creation property list in order to use the shuffle filter, and best results are usually obtained when the shuffle filter is set immediately prior to setting the deflate filter.

H5P.set_szip

Set compression method and compression options (since R2024b)

H5P.set_szip(dcplID,optionsMask,pixelsPerBlock) sets the compression method for the dataset creation property list identified by dcplID to H5Z_FILTER_SZIP.

 Details

Input Arguments

• dcplID — Dataset creation property list identifier.

• optionsMask — Encoding method for SZIP compression, specified as one of these string scalars or character vectors or their numeric equivalents:

  • "H5_SZIP_EC_OPTION_MASK" — Entropy encoding method. This method is best suited for data that has already been processed.

  • "H5_SZIP_NN_OPTION_MASK" — Nearest-neighbor encoding method. This method preprocesses the data and then applies the entropy encoding method.

• pixelsPerBlock — Number of pixels (HDF5 data elements) per block for SZIP compression, specified as an even integer from 2 to 32. The value of pixelsPerBlock must be less than or equal to the number of elements in each dataset chunk.

File Access Properties

H5P.get_alignment

Retrieve alignment properties

[threshold,alignment] = H5P.get_alignment(faplID) retrieves the current alignment properties from the file access property list specified by faplID.

H5P.get_driver

Low-level file driver

driverID = H5P.get_driver(plistID) returns the identifier of the low-level file driver associated with the file access property list or data transfer property list specified by plistID. See HDF5 documentation for a list of valid return values.

H5P.get_family_offset

Offset for family file driver

offset = H5P.get_family_offset(faplID) retrieves the value of offset from the file access property list specified by faplID. The offset output is the offset of the data in the HDF5 file that is stored on disk in the selected member file in a family of files.

H5P.get_fapl_core

Information about core file driver properties

[increment,backingStore] = H5P.get_fapl_core(faplID) queries the H5FD_CORE driver properties as set by H5P.set_fapl_core.

 Details

Output Arguments

• faplID — File access property list identifier.

• increment — Size in bytes of memory increments.

• backingStore — Boolean flag indicating whether to write the file contents to disk when the file is closed.

H5P.get_fapl_family

File access property list information

[membSize,membfaplID] = H5P.get_fapl_family(faplID) returns the size in bytes of each file member and the identifier of the file access property list for use with the family driver specified by faplID.

H5P.get_fapl_multi

Information about multifile access property list

[membmap,membfapl,membname,membaddr,relax] = H5P.get_fapl_multi(faplID) returns information about the multifile access property list specified by faplID.

 Details

Output Arguments

• membmap — Map of memory usage types to other memory usage types.

• membfapl — Property list for each memory usage type.

• membname — Name generator for names of member files.

• membaddr — Offsets within the virtual address space, returned as a value between 0 and "HADDR_MAX".

• relax — Boolean value that indicates whether to allow read-only access to incomplete file sets.

H5P.get_fclose_degree

File close degree

fcdegree = H5P.get_fclose_degree(faplID) returns the current setting of the file close degree property fcdegree in the file access property list specified by faplID.

 Details

Possible return values of fcdegree:

• "H5F_CLOSE_DEFAULT"

• "H5F_CLOSE_WEAK"

• "H5F_CLOSE_SEMI"

• "H5F_CLOSE_STRONG"

H5P.get_gc_references

Garbage collection references setting

tf = H5P.get_gc_references(faplID) returns a value of 1 if the garbage collection references property is on for the file access property list specified by faplID, and 0 if it is off.

H5P.get_libver_bounds

Library version bounds settings

[low,high] = H5P.get_libver_bounds(faplID) retrieves the lower and upper bounds, low and high, on the HDF5 library release versions that indirectly determine the object format versions used when creating objects in the file. This property is retrieved from the file access property list specified by the parameter faplID.

The H5P.get_libver_bounds function returns low and high as any of these enumerated values:

• "H5F_LIBVER_EARLIEST"

• "H5F_LIBVER_V18"

• "H5F_LIBVER_V110"

• "H5F_LIBVER_LATEST"

H5P.get_mdc_config

Metadata cache configuration

config = H5P.get_mdc_config(faplID) returns the current metadata cache configuration from the file access property list specified by faplID.

H5P.get_meta_block_size

Metadata block size setting

sz = H5P.get_meta_block_size(faplID) returns the current minimum size in bytes of new metadata block allocations.

H5P.get_multi_type

Type of data property for multifile driver

type = H5P.get_multi_type(plistID) returns the data setting from the file access or data transfer property list, plistID.

This function should only be used with an HDF5 file written as a set of files with the multifile driver.

H5P.get_sieve_buf_size

Maximum data sieve buffer size

sz = H5P.get_sieve_buf_size(faplID) returns the current maximum size of the data sieve buffer.

H5P.get_small_data_block_size

Small data block size setting

sz = H5P.get_small_data_block_size(faplID) returns the current setting for the size of the small data block of the file access property list specified by faplID.

H5P.set_alignment

Set alignment properties for file access property list

H5P.set_alignment(faplID,threshold,alignment) sets the alignment properties of the file access property list specified by faplID so that any file object greater than or equal in size to threshold, in bytes, is aligned on an address which is a multiple of the value specified by alignment.

In most cases, the default values of threshold and alignment result in the best performance.

H5P.set_family_offset

Set offset property for family of files

H5P.set_family_offset(faplID,offset)sets offset property in the file access property list specified by faplID for low-level access to a file in a family of files.

 Details

Input Arguments

• faplID — File access property list identifier.

• offset — User-determined location from the beginning of the HDF5 file in bytes.

H5P.set_fapl_core

Modify file access to use H5FD_CORE driver

H5P.set_fapl_core(faplID,increment,backingstore) modifies the file access property list specified by faplID to use the H5FD_CORE driver.

 Details

Input Arguments

• faplID — File access property list identifier.

• increment — Increment by which allocated memory is to be increased each time more memory is required.

• backingstore — Boolean flag that, when non-zero, indicates the file contents should be written to disk when the file is closed.

H5P.set_fapl_family

Set file access to use family driver

H5P.set_fapl_family(faplID,membsize,membfaplID) sets the file access property list, faplID, to use the family driver.

 Details

Input Arguments

• faplID — File access property list identifier.

• membsize — Size in bytes of each file member.

• membfaplID — Identifier of the file access property list to be used for each family member.

H5P.set_fapl_log

Set use of logging driver

H5P.set_fapl_log(faplID,logfile,flags,bufsize) modifies the file access property list, faplID, to use the logging driver H5FD_LOG.

 Details

Input Arguments

• faplID — File access property list identifier.

• logfile — Name of the file in which the logging entries are to be recorded.

• flags — Bit mask that specifies the types of activity to log. See the HDF Group HDF5 documentation for a list of available flag settings.

• bufsize — Size of the logging buffer.

H5P.set_fapl_multi

Set use of multifile driver

H5P.set_fapl_multi(faplID,relax) sets the file access property list, faplID, to access HDF5 files created with the multi-driver with default values provided by the HDF5 library. Specify relax as a value of 1 to allow read-only access to incomplete file sets.

H5P.set_fapl_multi(faplID,membmap,membfapl,membname,membaddr,relax) sets the file access property list to use the multifile driver with additional parameters.

 Details

Input Arguments

• faplID — File access property list identifier.

• membmap — Map of memory usage types to other memory usage types.

• membfapl — Property list for each memory usage type.

• membname — Name generator for names of member files.

• membaddr — Offset within the virtual address space at which each type of data storage begins.

• relax — Boolean value, specified as a value of 1 to allow read-only access to incomplete file sets, and 0 to disable it.

H5P.set_fapl_sec2

Set file access for sec2 driver

H5P.set_fapl_sec2(faplID) modifies the file access property list, faplID, to use the "H5FD_SEC2" driver.

H5P.set_fapl_split

Set file access for emulation of split file driver

H5P.set_fapl_split(faplID,metaext,metaplistID,rawext,rawplistID) is a compatibility function that enables the multifile driver to emulate the split driver from HDF5 Releases 1.0 and 1.2.

 Details

Input Arguments

• faplID — File access property list identifier.

• metaext — String scalar or character vector containing the metadata file name extension.

• metaplistID — File access property list identifier for the metadata file.

• rawext — String scalar or character vector containing the raw data file name extension.

• rawplistID — File access property list identifier for the raw data file.

H5P.set_fapl_stdio

Set file access for standard I/O driver

H5P.set_fapl_stdio(faplID) modifies the file access property list, faplID, to use the standard I/O driver, H5FD_STDIO.

H5P.set_fclose_degree

Set file access for file close degree

H5P.set_fclose_degree(faplID,degree) sets the file close degree property in the file access property list, faplID, to the value specified by degree.

 Details

Specify degree as one of these string scalars or character vectors:

• "H5F_CLOSE_WEAK"

• "H5F_CLOSE_SEMI"

• "H5F_CLOSE_STRONG"

• "H5F_CLOSE_DEFAULT".

H5P.set_gc_references

Set garbage collection references flag

H5P.set_gc_references(faplID,gcref) sets the flag for garbage collecting references for the file the file access property list identifier, faplID. Specify the flag gcref as a value of 1 to enable the garbage collection references property, and 0 to disable it.

H5P.set_libver_bounds

Set library version bounds for objects

H5P.set_libver_bounds(faplID,low,high) controls the range of library release versions that will be used when creating objects in a file. The object format versions are determined indirectly from the library release versions specified in the call.

This property is set in the file access property list specified by the parameter faplID.

 Details

Input Arguments

• faplID — File access property list identifier.

• low — Earliest possible format versions that the library will use when creating objects in the file. Note that earliest possible is different from earliest, as some features introduced in library versions later than 1.0.0 resulted in updates to object formats.

• high — Latest format versions that the library will be allowed to use when creating objects in the file.

Specify low and high as one of these enumerated values:

• "H5F_LIBVER_EARLIEST"

• "H5F_LIBVER_V18"

• "H5F_LIBVER_V110"

• "H5F_LIBVER_LATEST"

More About

The library supports multiple pairs of low and high. For more information on allowed library version bound combinations, see the H5P_SET_LIBVER_BOUNDS function on The HDF Group website.

H5P.set_mdc_config

Set initial metadata cache configuration

H5P.set_mdc_config(faplID,config) sets the initial metadata cache configuration in the file access property list specified by faplID to the supplied values in config. Before using this function, you should retrieve the current configuration using the H5P.get_mdc_config function.

Many of the fields in the config structure are intended to be used only in close consultation with the HDF Group.

H5P.set_meta_block_size

Set minimum metadata block size

H5P.set_meta_block_size(faplID,size) sets the minimum metadata block size size, in bytes, for the file access property list specified by faplID.

H5P.set_multi_type

Specify type of data to access with MULTI driver

H5P.set_multi_type(faplID,type) sets the type of data to access in the file access or data transfer property list specified by faplID.

 Details

Input Arguments

• faplID — File access property list identifier.

• type — Type of data property, specified as one of these values:

  • "H5FD_MEM_SUPER"

  • "H5FD_MEM_BTREE"

  • "H5FD_MEM_DRAW"

  • "H5FD_MEM_GHEAP"

  • "H5FD_MEM_LHEAP"

  • "H5FD_MEM_OHDR"

H5P.set_sieve_buf_size

Set maximum size of data sieve buffer

H5P.set_sieve_buf_size(faplID,bufsize) sets the maximum size in bytes of the data sieve buffer, bufsize, which is used by the file drivers that can sieve data.

H5P.set_small_data_block_size

Set size of block reserved for small data

H5P.set_small_data_block_size(faplID,size) sets the maximum size, in bytes, of a contiguous block reserved for small data in the file access property list specified by faplID.

File Creation Properties

H5P.get_istore_k

1/2 rank of indexed storage B-tree

ik = H5P.get_istore_k(fcplID) returns the chunked storage B-tree 1/2 rank of the file creation property list specified by fcplID.

H5P.get_sizes

Size of offsets and lengths

[offset,lengths] = H5P.get_sizes(fcplID) returns the size of the offsets and lengths used of an HDF5 file.

H5P.get_sym_k

Size of B-tree 1/2 rank and leaf node 1/2 size

[ik,lk] = H5P.get_sym_k(fcplID) returns the size of the symbol table B-tree 1/2 rank, ik, and the symbol table leaf node 1/2 size, lk.

H5P.get_userblock

Size of user block

sz = H5P.get_userblock(fcplID) returns the size of a user block.

H5P.get_version

Version information for file creation property list

[superblock,freelist,symtable,shhdr] = H5P.get_version(fcplID) returns the version of the super block, the global freelist, the symbol table, and the shared object header. Retrieving this information requires the file creation property list.

H5P.set_istore_k

Set size of parameter for indexing chunked datasets

H5P.set_istore_k(fcplID,ik) sets the size of the parameter used to control the B-trees for indexing chunked datasets in the file creation property list specified by fcplID. The ik argument is one half the rank of a tree that stores chunked raw data.

H5P.set_sizes

Set byte size of offsets and lengths

H5P.set_sizes(fcplID,offset,lengths) sets the byte size of the offsets and lengths used to address objects in an HDF5 file.

H5P.set_sym_k

Set size of parameters used to control symbol table nodes

H5P.set_sym_k(faplID,ik,lk) sets the size of parameters used to control the symbol table nodes for the file access property list specified by faplID. The ik argument is one half the rank of a tree that stores a symbol table for a group, and lk is one half of the number of symbols that can be stored in a symbol table node.

H5P.set_userblock

Set the user block size

H5P.set_userblock(fcplID,size) sets the user block size of the file creation property list specified by fcplID.

Object Copy and Object Creation Properties

H5P.get_attr_creation_order

Tracking order and indexing settings

crtOrder = H5P.get_attr_creation_order(ocplID) retrieves tracking and indexing settings for attribute creation order.

 Details

The H5P.get_attr_creation_order function returns crtOrder as one or both of these values:

• "H5P_CRT_ORDER_TRACKED" — Attribute creation order is tracked.

• "H5P_CRT_ORDER_INDEXED" — Attribute creation order is indexed.

• 0 — Attribute creation order is neither tracked nor indexed.

H5P.get_attr_phase_change

Retrieve attribute phase change thresholds

[maxCompact,minDense] = H5P.get_attr_phase_change(ocplID) returns attribute phase change thresholds for the dataset or group with creation property list specified by ocplID.

 Details

Output Arguments

• maxCompact — Maximum number of attributes to be stored in compact storage. The default maximum number of attributes is 8.

• minDense — Minimum number of attributes to be stored in dense storage. The default minimum number of attributes is 6.

H5P.get_copy_object

Properties to be used when object is copied

options = H5P.get_copy_object(ocplID) retrieves the properties currently specified in the object copy property list identifier ocplID, which will be invoked when a new copy is made of an existing object.

H5P.set_attr_creation_order

Set tracking of attribute creation order

H5P.set_attr_creation_order(ocplID,flags) sets tracking and indexing of attribute creation order. By default, the attribute creation order is neither tracked nor indexed.

 Details

he creation order flags should be either of these values or a bitwise-or of them:

• "H5P_CRT_ORDER_TRACKED" — Attribute creation order is tracked.

• "H5P_CRT_ORDER_INDEXED" — Attribute creation order is indexed.

H5P.set_attr_phase_change

Set attribute storage phase change thresholds

H5P.set_attr_phase_change(ocplID,maxCompact,minDense) sets the attribute storage phase change thresholds for the group or dataset with creation order property list specified by ocplID.

 Details

Input Arguments

• ocplID — Creation order property list identifier.

• maxCompact — Maximum number of attributes to be stored in compact storage. The default maximum number of attributes is 8.

• minDense — Minimum number of attributes to be stored in dense storage. The default minimum number of attributes is 6.

H5P.set_copy_object

Set properties to be used when objects are copied

H5P.set_copy_object(ocplID,options) sets the properties in the object copy property list ocplID that will be invoked when a new copy is made of an existing object.

ocplID is the object copy property list and specifies the properties governing the copying of the object.

 Details

Specify optionsas a string scalar or character vector containing one of these flag names, or its equivalent numerical value:

• "H5O_COPY_SHALLOW_HIERARCHY_FLAG" — Copy only immediate members of a group. If you do not specify this flag, MATLAB recursively copies all objects below the group.

• "H5O_COPY_EXPAND_SOFT_LINK_FLAG" — Expand soft links into new objects. If you do not specify this flag, MATLAB keeps soft links as they are.

• "H5O_COPY_EXPAND_EXT_LINK_FLAG" — Expand external link into new objects. If you do not specify this flag, MATLAB keeps external links as they are.

• "H5O_COPY_EXPAND_REFERENCE_FLAG" — Copy objects that are pointed to by references. If you do not specify this flag, MATLAB updates only the values of object references.

• "H5O_COPY_WITHOUT_ATTR_FLAG" — Copy object without copying attributes. If you do not specify this flag, MATLAB copies object along with all its attributes.

• "H5O_COPY_MERGE_COMMITTED_DTYPE_FLAG" — Use a matching committed datatype in the destination file when copying a committed datatype, a dataset with a committed datatype, or an object with an attribute of committed datatype.

Group Creation Properties

H5P.get_create_intermediate_group

Determine creation of intermediate groups

output = H5P.get_create_intermediate_group(lcplID) returns a positive value if the link creation property list lcplID is set to enable creating missing intermediate groups, and in this case missing intermediate groups will be created; otherwise, H5P.get_create_intermediate_group returns 0, and missing intermediate groups will not be created.

H5P.get_link_creation_order

Query if link creation order is tracked

crtOrder = H5P.get_link_creation_order(gcplID) queries whether link creation order is tracked or indexed in a group with creation property list identifiergcplID.

 Details

The H5P.get_link_creation_order function returns crtOrder as one or both of these constant values:

• "H5P_CRT_ORDER_TRACKED" — Attribute creation order is tracked.

• "H5P_CRT_ORDER_INDEXED" — Attribute creation order is indexed.

H5P.get_link_phase_change

Query settings for conversion between groups

[maxCompact,minDense] = H5P.get_link_phase_change(gcplID) retrieves the settings for conversion between compact and dense groups.

 Details

Output Arguments

• maxCompact — Maximum number of links to store as header messages in the group header before converting the group to the dense format. Groups that are in the compact format and exceed this number of links are automatically converted to the dense format.

• minDense — Minimum number of links to store in the dense format. Groups which are in dense format and in which the number of links falls below this number are automatically converted back to the compact format.

H5P.set_create_intermediate_group

Set creation of intermediate groups

H5P.set_create_intermediate_group(lcplID,flag) specifies in the link creation property list lcplID whether to create missing intermediate groups.

H5P.set_link_creation_order

Set creation order tracking and indexing

H5P.set_link_creation_order(gcplID,crtOrder) sets creation order tracking and indexing for links in the group with group creation property list identifier gcplID.

 Details

Specify the creation order flags as one or both of these constant values:

• "H5P_CRT_ORDER_TRACKED" — Attribute creation order is tracked.

• "H5P_CRT_ORDER_INDEXED" — Attribute creation order is indexed.

If only "H5P_CRT_ORDER_TRACKED" is set, HDF5 will track link creation order in any group created with the group creation property list gcplID. If both "H5P_CRT_ORDER_TRACKED" and "H5P_CRT_ORDER_INDEXED" are set, HDF5 will track link creation order in the group and index links on that property.

H5P.set_link_phase_change

Set parameters for group conversion

H5P.set_link_phase_change(gcplID,maxCompact,minDense) sets the parameters for conversion between compact and dense groups.

 Details

Input Arguments

• gcplID — Group creation property list identifier.

• maxCompact — Maximum number of links to store as header messages in the group header before converting the group to the dense format. Groups that are in the compact format and exceed this number of links are automatically converted to the dense format.

• minDense — Minimum number of links to store in the dense format. Groups which are in dense format and in which the number of links falls below this number are automatically converted back to the compact format.

HDF5 String Properties

H5P.get_char_encoding

Character encoding

encoding = H5P.get_char_encoding(plistID) returns the character encoding used to encode strings or object names that are created with the property list specified by plistID.

 Details

H5P.get_char_encoding returns encoding as one of these values:

• "H5T_CSET_UTF8" — Encode strings or object names using UTF-8 encoding.

• "H5T_CSET_ASCII" — Encode strings or object names using ASCII encoding.

H5P.set_char_encoding

Set character encoding used to encode strings

H5P.set_char_encoding(plistID,encoding) sets the character encoding used to encode strings or object names that are created with the property list specified by plistID.

 Details

Specify encoding as one of these values:

• "H5T_CSET_UTF8" — Use UTF-8 encoding to encode strings or object names.

• "H5T_CSET_ASCII" — Use ASCII encoding to encode strings or object names.

Virtual Dataset (VDS) Properties

H5P.get_virtual_count

Number of mappings for the virtual dataset

count = H5P.get_virtual_count(dcplID) gets the number of mappings for the virtual dataset count associated with the dataset creation property list identifier dcplID.

H5P.get_virtual_dsetname

Name of source dataset

srcname = H5P.get_virtual_dsetname(dcplID,mapIdx) returns the name of the source dataset srcname used in the mapping for the virtual dataset.

 Details

Input Arguments

• dcplID — Identifier of the virtual dataset creation property list.

• mapIdx — A numeric value greater than or equal to zero and less than the number of mappings returned by the H5P.get_virtual_count function.

H5P.get_virtual_filename

Name of file for source dataset

srcname = H5P.get_virtual_filename(dcplID,mapIdx) returns the name of the file for a source dataset used in the mapping for the virtual dataset.

 Details

Input Arguments

• dcplID — Identifier of the virtual dataset creation property list.

• mapIdx — A numeric value greater than or equal to zero and less than the number of mappings returned by the H5P.get_virtual_count function.

H5P.get_virtual_printf_gap

Maximum number of missing source files or datasets with printf-style names

gapsize = H5P.get_virtual_printf_gap(daplID) returns the maximum number of printf-style files and/or datasets allowed to be missing for determining the extent of an unlimited virtual dataset with printf-style mappings. The default library value for gapsize is 0.

H5P.get_virtual_srcspace

Dataspace identifier for source dataset

srcdspaceID = H5P.get_virtual_srcspace(dcplID,mapIdx) returns a dataspace identifier for the selection within the source dataset used in the mapping.

 Details

Input Arguments

• dcplID — Identifier of the virtual dataset creation property list.

• mapIdx — A numeric value greater than or equal to zero and less than the number of mappings returned by the H5P.get_virtual_count function.

H5P.get_virtual_view

View of virtual dataset

view = H5P.get_virtual_view(daplID) returns the view of a dataset access property list identifier for the virtual dataset daplID.

 Details

H5P.get_virtual_view returns view as one of these flags:

• "H5D_VDS_FIRST_MISSING"

• "H5D_VDS_LAST_AVAILABLE"

H5P.get_virtual_vspace

Dataspace identifier for selection

vdspaceID = H5P.get_virtual_vspace(dcplID,mapIdx) returns a dataspace identifier for the selection within the virtual dataset used in the mapping. The dataspace identifier must be closed with H5S.close().

 Details

Input Arguments

• dcplID — Identifier of the virtual dataset creation property list.

• mapIdx — A numeric value greater than or equal to zero and less than the number of mappings returned by the H5P.get_virtual_count function.

H5P.set_virtual

Set the mapping between virtual and source datasets

H5P.set_virtual(dcplID,vspaceID,srcfilename,srcdsetname,srcspaceID) maps the elements of the virtual dataset, described by the virtual dataspace identifier vspaceID, to the elements of the source dataset described by the source dataspace identifier srcspaceID.

 Details

Input Arguments

• dcplID — Identifier of the virtual dataset creation property list to use when creating the virtual dataset.

• vspaceID — Dataspace identifier with the selection within the virtual dataset applied. The selection within the virtual dataset can be of unlimited dimension.

• srcfilename — Name of the HDF5 file where the source dataset is located.

• srcdsetname — Path to the HDF5 dataset in the file specified by srcfilename.

• srcspaceID — Dataspace identifier of the source dataset with a selection applied.

H5P.set_virtual_printf_gap

Set maximum number of missing source files or datasets with printf-style names

H5P.set_virtual_printf_gap(daplID,gapsize) sets the access property list for the virtual dataset daplID to instruct the library to stop looking for the mapped data stored in the files and/or datasets with the printf-style names after not finding gapsize files and/or datasets. The found source files and datasets will determine the extent of the unlimited virtual dataset with the printf-style mappings.

 Details

Input Arguments

• daplID — Identifier of the virtual dataset access property list.

• gapsize — Maximum number of files or datasets allowed to be missing for determining the extent of an unlimited virtual dataset with printf-style mappings. The default value of gapsize is 0.

H5P.set_virtual_view

Set view of virtual dataset

H5P.set_virtual_view(daplID,view) sets the view of the virtual dataset to include or exclude missing mapped elements.

 Details

Input Arguments

• daplID — Identifier of the virtual dataset access property list.

• view — Flag specifying the extent of the data to be included in the view. Specify view as one of these values:

  • "H5D_VDS_FIRST_MISSING" — View includes all data before the first missing mapped data.

  • "H5D_VDS_LAST_AVAILABLE" — View includes all available mapped data.

Metadata Cache Fine-Tuning Properties

H5P.get_metadata_read_attempts

Number of read attempts

attempts = H5P.get_metadata_read_attempts(faplID) returns the number of read attempts from a file access property list specified by faplID.

 Details

Note

The number of read attempts used by the library will depend on how the file is opened and whether the user sets the number of read attempts via this routine:

For a file opened with SWMR access:

• If the user sets the number of attempts to N, the library will use N.

• If the user does not set the number of attempts, the library will use the default for SWMR access (100).

For a file opened with non-SWMR access, the library will always use the default for non-SWMR access (1). The value set via this routine does not have any effect during non-SWMR access.

H5P.set_metadata_read_attempts

Set number of read attempts

H5P.set_metadata_read_attempts(faplID,attempts) sets the number of reads that the library will try when reading checksummed metadata in an HDF5 file opened with SWMR access. Specify attempts as a numeric value greater than zero.

Partial Edge Chunk Properties

H5P.get_chunk_opts

Edge chunk option setting

options = H5P.get_chunk_opts(dcplID) returns the edge chunk option setting stored in the dataset creation property list specified by dcplID.

H5P.set_chunk_opts

Set edge chunk option setting

H5P.set_chunk_opts(dcplID,options) specifies storage options for chunks on the edge of a dataset's dataspace for dataset creation property list. This allows for performance tuning in cases where the dataset size may not be a multiple of the chunk size and the handling of partial edge chunks can impact performance.

 Details

Input Arguments

• dcplID — Dataset creation property list identifier.

• options — Edge chunk option flag. Specify options as one of these values:

  • "H5D_CHUNK_DONT_FILTER_PARTIAL_CHUNKS" — Partial edge chunks are not filtered. Datasets created with this flag enabled will be inaccessible with HDF5 library versions before Release 1.10.

  • 0 — Filter and compress partial edge chunks.

Examples

Query Information About Property Lists and Datasets

Query information about property lists associated with the file "example.h5" and some datasets in the file.

Open the file and query the property lists.

fid = H5F.open("example.h5");
fapl = H5F.get_access_plist(fid);
fcpl = H5F.get_create_plist(fid);

Get the library version bounds.

[low,high] = H5P.get_libver_bounds(fapl);

Determine whether two property lists are equal.

if H5P.equal(fapl,fcpl)
    fprintf("property lists are equal\n");
else
    fprintf("property lists are not equal\n");
end

Determine whether a property in the property list exists.

if H5P.exist(fapl,"sieve_buf_size")
    fprintf("sieve buffer size property exists\n");
else
    fprintf("sieve buffer size property does not exist\n");
end

Return the size of chunks in a dataset.

dsID = H5D.open(fid,"/g4/time");
dcpl = H5D.get_create_plist(dsID);
[rank,chunkDims] = H5P.get_chunk(dcpl);
H5D.close(dsID);

Determine the layout of a dataset.

dsID = H5D.open(fid,"/g3/integer");
dcpl = H5D.get_create_plist(dsID);
layout = H5P.get_layout(dcpl);
switch(layout)
    case H5ML.get_constant_value("H5D_COMPACT")
        fprintf("layout is compact\n");
    case H5ML.get_constant_value("H5D_CONTIGUOUS")
        fprintf("layout is contiguous\n");
    case H5ML.get_constant_value("H5D_CHUNKED")
        fprintf("layout is chunked\n");
    case H5ML.get_constant_value("H5D_VIRTUAL")
        fprintf("layout is virtual\n");
end
H5D.close(dsID);
H5F.close(fid);

Create Dataset with Chunking and Compression

Set the chunk size and compression settings for dataset creation.

Create new property lists.

fcpl = H5P.create("H5P_FILE_CREATE");
fapl = H5P.create("H5P_FILE_ACCESS");

Set the library version bounds, create the HDF5 file, and create a copy of its datatype.

H5P.set_libver_bounds(fapl,"H5F_LIBVER_LATEST","H5F_LIBVER_LATEST");
fid = H5F.create("myfile.h5","H5F_ACC_TRUNC",fcpl,fapl);
typeID = H5T.copy("H5T_NATIVE_DOUBLE");

Create a simple dataspace with the upper limit on the size of each dimension as unlimited.

unlimited = H5ML.get_constant_value("H5S_UNLIMITED");
dims = [512 1024];
h5_dims = fliplr(dims);
h5_maxdims = [unlimited unlimited];
spaceID = H5S.create_simple(2,h5_dims,h5_maxdims);
dcpl = H5P.create("H5P_DATASET_CREATE");

Set the chunk size.

chunkDims = [512 1024];
h5_chunkDims = fliplr(chunkDims);
H5P.set_chunk(dcpl,h5_chunkDims);

Set the compression method to "H5D_COMPRESS_DEFLATE" and the level to 5. Create the dataset.

H5P.set_deflate(dcpl,5);
dsID1 = H5D.create(fid,"DS1",typeID,spaceID,dcpl);

Close the identifiers.

H5D.close(dsID1);
H5P.close(dcpl);
H5S.close(spaceID);
H5T.close(typeID);
H5F.close(fid);
H5P.close(fapl);
H5P.close(fcpl);

Create Dataset with Specific Storage Layout

Create a dataset and set the storage type for dataset creation.

Create new property lists.

fcpl = H5P.create("H5P_FILE_CREATE");
fapl = H5P.create("H5P_FILE_ACCESS");

Set the library version bounds, create the HDF5 file, and create a copy of its datatype.

H5P.set_libver_bounds(fapl,"H5F_LIBVER_LATEST","H5F_LIBVER_LATEST");
fid = H5F.create("myfile.h5","H5F_ACC_TRUNC",fcpl,fapl);
typeID = H5T.copy("H5T_NATIVE_DOUBLE");

Create a simple dataspace.

dims = [100 200];
h5_dims = fliplr(dims);
spaceID = H5S.create_simple(2,h5_dims,[]);

dcpl = H5P.create("H5P_DATASET_CREATE");
layout = H5ML.get_constant_value("H5D_CONTIGUOUS");

Set the storage layout of the dataset, then create the dataset.

H5P.set_layout(dcpl,layout);
dsID2 = H5D.create(fid,"DS2",typeID,spaceID,dcpl);

Close all identifiers.

H5D.close(dsID2);
H5P.close(dcpl);
H5S.close(spaceID);
H5T.close(typeID);
H5F.close(fid);
H5P.close(fapl);
H5P.close(fcpl);

Set Character Encoding and Get information on Properties

Set the character encoding for a property list, get information about the filters of a dataset, and iterate over properties in its property list.

Set the character encoding and retrieve it.

plist1 = H5P.create("H5P_ATTRIBUTE_CREATE");
H5P.set_char_encoding(plist1,"H5T_CSET_UTF8");
actValue1 = H5P.get_char_encoding(plist1);
H5P.close(plist1);

Return information about the specified filter, and set the N-Bit filter, H5Z_FILTER_NBIT, in dcpl.

filterID = H5ML.get_constant_value("H5Z_FILTER_NBIT");
dcpl = H5P.create("H5P_DATASET_CREATE");
H5P.set_nbit(dcpl);
[~,~,actName] = H5P.get_filter_by_id(dcpl,filterID);
H5P.close(dcpl);

Iterate over properties in the property list and invoke a callback function whose handle is @propertyIterator. Then, close the identifier.

fid = H5F.open("example.h5");
dsID = H5D.open(fid,"/g1/g1.1/dset1.1.1");
plist = H5D.get_create_plist(dsID);
[output,idx] = H5P.iterate(plist,0,@propertyIterator);
H5D.close(dsID);
H5F.close(fid);

The callback function propertyIterator is defined as the following:

function output = propertyIterator(~,prop_name)
fprintf("%s\n",prop_name);
output = 0;

Version History

Introduced before R2006a

R2024b: Work with datasets with SZIP compression

You can use the H5P.set_szip function to enable SZIP compression on datasets.

These functions accept the value "H5Z_FILTER_SZIP" for the filter input argument:

• H5P.get_filter_by_id

• H5P.modify_filter

• H5P.remove_filter

• H5P.set_filter

R2021b: Work with Virtual Datasets (VDS)

• You can use the layout "H5D_VIRTUAL" with H5P.set_layout to set the storage to draw raw data from multiple datasets in different files, and you can get this layout with H5P.get_layout.

• You can use the following functions to get information about virtual datasets.

  • Use H5P.get_virtual_count to get the number of mappings.

  • Use H5P.get_virtual_dsetname to get the name of the source dataset in the mapping.

  • Use H5P.get_virtual_filename to get the name of a file.

  • Use H5P.get_virtual_src_space to get the dataspace identifier.

  • Use H5P.get_virtual_vspace to get a virtual dataspace identifier for selection.

• You can use H5P.get_virtual_printf_gap and H5P.set_virtual_printf_gap to get and set, respectively, the maximum number of missing source files or datasets with printf-style names.

• You can use H5P.get_virtual_view and H5P.set_virtual_view to get and set, respectively, a view of a virtual dataset.

• You can use H5P.set_virtual to set the mapping between virtual and source datasets.

R2021b: Fine-tune the metadata cache by controlling read attempts

You can use H5P.get_metadata_read_attempts and H5P.set_metadata_read_attempts to get and set, respectively, the number of read attempts.

R2021b: Work with partial edge chunk settings

You can use H5P.get_chunk_opts and H5P.set_chunk_opts to get and set, respectively, the edge chunk option setting.