hdfread (MATLAB Functions)

Extract data from an HDF or HDF-EOS file

Syntax

data = hdfread(filename, dataset)
data = hdfread(hinfo)
data = hdfread(...,param1,value1,param2,value2,...)
[data,map] = hdfread(...)

Description

data = hdfread(filename, dataset) returns all the data in the specified data set dataset from the HDF or HDF-EOS file filename. To determine the names of the data sets in an HDF file, use the hdfinfo function. The information returned by hdfinfo contains structures describing the data sets contained in the file. You can extract one of these structures and pass it directly to hdfread.

Note hdfread can be used on Version 4.x HDF files or Version 2.x HDF-EOS files.

data = hdfread(hinfo) returns all the data in the data set specified in the structure hinfo. The hinfo structure can be extracted from the data returned by the hdfinfo function.

data = hdfread(...,param1,value1,param2,value2,...) returns subsets of the data according to the specified parameter and value pairs. See the tables below to find the valid parameters and values for different types of data sets.

[data,map] = hdfread(...) returns the image data and the colormap map for an 8-bit raster image.

Subsetting Parameters

The following tables show the subsetting parameters that can be used with the hdfread function for certain types of HDF data. These data types are

Note the following:

If a parameter requires multiple values, the values must be stored in a cell array. For example, the 'Index' parameter requires three values: start, stride, and edge. Enclose these values in curly braces as a cell array.

hdfread(dataset_name, 'Index', {start,stride,edge})

All values that are indices are 1-based.

Subsetting Parameters for HDF Scientific Data (SD) Data Sets

When you are working with HDF SD files, hdfread supports the parameters listed in this table.

Parameter
Description

'Index'
Three-element cell array, {start,stride,edge}, specifying the location, range, and values to be read from the data set
start -- A 1-based array specifying the position in the file to begin reading

Default: 1, start at the first element of each dimension. The values specified must not exceed the size of any dimension of the data set.

stride -- A 1-based array specifying the interval between the values to read

Default: 1, read every element of the data set.

edge -- A 1-based array specifying the length of each dimension to read

Default: An array containing the lengths of the corresponding dimensions

Parameter	Description
`'Index'`	Three-element cell array, `{start,stride,edge}`, specifying the location, range, and values to be read from the data set `start` -- A 1-based array specifying the position in the file to begin reading Default: `1`, start at the first element of each dimension. The values specified must not exceed the size of any dimension of the data set. `stride` -- A 1-based array specifying the interval between the values to read Default: `1`, read every element of the data set. `edge` -- A 1-based array specifying the length of each dimension to read Default: An array containing the lengths of the corresponding dimensions

For example, this code reads the data set Example SDS from the HDF file example.hdf. The 'Index' parameter specifies that hdfread start reading data at the beginning of each dimension, read until the end of each dimension, but only read every other data value in the first dimension.

hdfread('example.hdf','Example SDS', ...
                   'Index', {[], [2 1], []})

Subsetting Parameters for HDF Vdata Sets

When you are working with HDF Vdata files, hdfread supports these parameters.

Parameter
Description

'Fields'
Text string specifying the name of the data set field to be read from. When specifying multiple field names, use a comma-separated list.

'FirstRecord'
1-based number specifying the record from which to begin reading

'NumRecords'
Number specifying the total number of records to read

Parameter	Description
`'Fields'`	Text string specifying the name of the data set field to be read from. When specifying multiple field names, use a comma-separated list.
`'FirstRecord'`	1-based number specifying the record from which to begin reading
`'NumRecords'`	Number specifying the total number of records to read

For example, this code reads the Vdata set Example Vdata from the HDF file example.hdf.

hdfread('example.hdf', 'Example Vdata', 'FirstRecord', 400, 
'NumRecords', 50)

Subsetting Parameters for HDF-EOS Grid Data

When you are working with HDF-EOS grid data, hdfread supports three types of parameters:

Required parameters
Optional parameters

Mutually exclusive parameters -- You can only specify one of these parameters in a call to hdfread, and you cannot use these parameters in combination with any optional parameter.

Parameter
Description

Required Parameter

'Fields'
String naming the data set field to be read. You can specify only one field name for a Grid data set.

Mutually Exclusive Optional Parameters

'Index'
Three-element cell array, {start,stride,edge}, specifying the location, range, and values to be read from the data set
start -- An array specifying the position in the file to begin reading

Default: 1, start at the first element of each dimension. The values must not exceed the size of any dimension of the data set.

stride -- An array specifying the interval between the values to read

Default: 1, read every element of the data set.

edge -- An array specifying the length of each dimension to read

Default: An array containing the lengths of the corresponding dimensions

'Interpolate'
Two-element cell array, {longitude,latitude}, specifying the longitude and latitude points that define a region for bilinear interpolation. Each element is an N-length vector specifying longitude and latitude coordinates.

'Pixels'

Two-element cell array, {longitude,latitude}, specifying the longitude and latitude coordinates that define a region. Each element is an N-length vector specifying longitude and latitude coordinates. This region is converted into pixel rows and columns with the origin in the upper left corner of the grid.
Note: This is the pixel equivalent of reading a 'Box' region.

'Tile'
Vector specifying the coordinates of the tile to read, for HDF-EOS Grid files that support tiles

Optional Parameters

'Box'
Two-element cell array, {longitude,latitude}, specifying the longitude and latitude coordinates that define a region. longitude and latitude are each two-element vectors specifying longitude and latitude coordinates.

'Time'
Two-element cell array, [start stop], where start and stop are numbers that specify the start and end-point for a period of time

'Vertical'
Two-element cell array, {dimension, range}
dimension -- String specifying the name of the data set field to be read from. You can specify only one field name for a Grid data set.

range -- Two-element array specifying the minimum and maximum range for the subset. If dimension is a dimension name, then range specifies the range of elements to extract. If dimension is a field name, then range specifies the range of values to extract.

'Vertical' subsetting can be used alone or in conjunction with 'Box' or 'Time'. To subset a region along multiple dimensions, vertical subsetting can be used up to eight times in one call to hdfread.

Parameter	Description
Required Parameter
`'Fields'`	String naming the data set field to be read. You can specify only one field name for a Grid data set.
Mutually Exclusive Optional Parameters
`'Index'`	Three-element cell array, `{start,stride,edge}`, specifying the location, range, and values to be read from the data set `start` -- An array specifying the position in the file to begin reading Default: `1`, start at the first element of each dimension. The values must not exceed the size of any dimension of the data set. `stride` -- An array specifying the interval between the values to read Default: `1`, read every element of the data set. `edge` -- An array specifying the length of each dimension to read Default: An array containing the lengths of the corresponding dimensions
`'Interpolate'`	Two-element cell array, `{longitude,latitude}`, specifying the longitude and latitude points that define a region for bilinear interpolation. Each element is an N-length vector specifying longitude and latitude coordinates.
`'Pixels'`	Two-element cell array, `{longitude,latitude}`, specifying the longitude and latitude coordinates that define a region. Each element is an N-length vector specifying longitude and latitude coordinates. This region is converted into pixel rows and columns with the origin in the upper left corner of the grid. Note: This is the pixel equivalent of reading a `'Box'` region.
`'Tile'`	Vector specifying the coordinates of the tile to read, for HDF-EOS Grid files that support tiles
Optional Parameters
`'Box'`	Two-element cell array, `{longitude,latitude}`, specifying the longitude and latitude coordinates that define a region. `longitude` and `latitude` are each two-element vectors specifying longitude and latitude coordinates.
`'Time'`	Two-element cell array, `[start stop]`, where `start` and `stop` are numbers that specify the start and end-point for a period of time
`'Vertical'`	Two-element cell array, `{dimension, range}` `dimension` -- String specifying the name of the data set field to be read from. You can specify only one field name for a Grid data set. `range` -- Two-element array specifying the minimum and maximum range for the subset. If `dimension` is a dimension name, then `range` specifies the range of elements to extract. If `dimension` is a field name, then `range` specifies the range of values to extract. `'Vertical'` subsetting can be used alone or in conjunction with `'Box'` or `'Time'`. To subset a region along multiple dimensions, vertical subsetting can be used up to eight times in one call to `hdfread`.

For example,

hdfread(grid_dataset, 'Fields', fieldname, ...
              'Vertical', {dimension, [min, max]})

Subsetting Parameters for HDF-EOS Point Data

When you are working with HDF-EOS Point data, hdfread has two required parameters and three optional parameters.

Parameter
Description

Required Parameters

'Fields'
String naming the data set field to be read. For multiple field names, use a comma-separated list.

'Level'
1-based number specifying which level to read from in an HDF-EOS Point data set

Optional Parameters

'Box'
Two-element cell array, {longitude,latitude}, specifying the longitude and latitude coordinates that define a region. longitude and latitude are each two-element vectors specifying longitude and latitude coordinates.

'RecordNumbers'
Vector specifying the record numbers to read

'Time'
Two-element cell array, [start stop], where start and stop are numbers that specify the start and endpoint for a period of time

Parameter	Description
Required Parameters
`'Fields'`	String naming the data set field to be read. For multiple field names, use a comma-separated list.
`'Level'`	1-based number specifying which level to read from in an HDF-EOS Point data set
Optional Parameters
`'Box'`	Two-element cell array, `{longitude,latitude}`, specifying the longitude and latitude coordinates that define a region. `longitude` and `latitude` are each two-element vectors specifying longitude and latitude coordinates.
`'RecordNumbers'`	Vector specifying the record numbers to read
`'Time'`	Two-element cell array, `[start stop]`, where `start` and `stop` are numbers that specify the start and endpoint for a period of time

For example,

hdfread(point_dataset, 'Fields', {field1, field2}, ...
'Level', level, 'RecordNumbers', [1:50, 200:250])

Subsetting Parameters for HDF-EOS Swath Data

When you are working with HDF-EOS Swath data, hdfread supports three types of parameters:

Required parameters
Optional parameters
Mutually exclusive

You can only use one of the mutually exclusive parameters in a call to hdfread, and you cannot use these parameters in combination with any optional parameter.

Parameter
Description

Required Parameter

'Fields'
String naming the data set field to be read. You can specify only one field name for a Swath data set.

Mutually Exclusive Optional Parameters

'Index'
Three-element cell array, {start,stride,edge}, specifying the location, range, and values to be read from the data set
start -- An array specifying the position in the file to begin reading

Default: 1, start at the first element of each dimension. The values must not exceed the size of any dimension of the data set.

stride -- An array specifying the interval between the values to read

Default: 1, read every element of the data set.

edge -- An array specifying the length of each dimension to read

Default: An array containing the lengths of the corresponding dimensions

'Time'
Three-element cell array, {start, stop, mode}, where start and stop specify the beginning and the endpoint for a period of time, and mode is a string defining the criterion for the inclusion of a cross track in a region. The cross track is within a region if any of these conditions is met:
Its midpoint is within the box (mode='midpoint').

Either endpoint is within the box (mode='endpoint').

Any point is within the box (mode='anypoint').

Optional Parameters

'Box'
Three-element cell array, {longitude, latitude, mode} specifying the longitude and latitude coordinates that define a region. longitude and latitude are two-element vectors that specify longitude and latitude coordinates. mode is a string defining the criterion for the inclusion of a cross track in a region. The cross track is within a region if any of these conditions is met:
Its midpoint is within the box (mode='midpoint').

Either endpoint is within the box (mode='endpoint').

Any point is within the box (mode='anypoint').

'ExtMode'

String specifying whether geolocation fields and data fields must be in the same swath (mode='internal'), or can be in different swaths (mode='external')
Note: mode is only used when extracting a time period or a region.

'Vertical'
Two-element cell array, {dimension, range}
dimension is a string specifying either a dimension name or field name to subset the data by.

range is a two-element vector specifying the minimum and maximum range for the subset. If dimension is a dimension name, then range specifies the range of elements to extract. If dimension is a field name, then range specifies the range of values to extract.

'Vertical' subsetting can be used alone or in conjunction with 'Box' or 'Time'. To subset a region along multiple dimensions, vertical subsetting can be used up to eight times in one call to hdfread.

Parameter	Description
Required Parameter
`'Fields'`	String naming the data set field to be read. You can specify only one field name for a Swath data set.
Mutually Exclusive Optional Parameters
`'Index'`	Three-element cell array, `{start,stride,edge}`, specifying the location, range, and values to be read from the data set `start` -- An array specifying the position in the file to begin reading Default: `1`, start at the first element of each dimension. The values must not exceed the size of any dimension of the data set. `stride` -- An array specifying the interval between the values to read Default: `1`, read every element of the data set. `edge` -- An array specifying the length of each dimension to read Default: An array containing the lengths of the corresponding dimensions
`'Time'`	Three-element cell array, `{start, stop, mode}`, where `start` and `stop` specify the beginning and the endpoint for a period of time, and `mode` is a string defining the criterion for the inclusion of a cross track in a region. The cross track is within a region if any of these conditions is met: Its midpoint is within the box (`mode='midpoint'`). Either endpoint is within the box (`mode='endpoint'`). Any point is within the box (`mode='anypoint'`).
Optional Parameters
`'Box'`	Three-element cell array, `{longitude, latitude, mode}` specifying the longitude and latitude coordinates that define a region. `longitude` and `latitude` are two-element vectors that specify longitude and latitude coordinates. `mode` is a string defining the criterion for the inclusion of a cross track in a region. The cross track is within a region if any of these conditions is met: Its midpoint is within the box (`mode='midpoint'`). Either endpoint is within the box (`mode='endpoint'`). Any point is within the box (`mode='anypoint'`).
`'ExtMode'`	String specifying whether geolocation fields and data fields must be in the same swath (`mode='internal'`), or can be in different swaths (`mode='external'`) Note: `mode` is only used when extracting a time period or a region.
`'Vertical'`	Two-element cell array, `{dimension, range}` `dimension` is a string specifying either a dimension name or field name to subset the data by. `range` is a two-element vector specifying the minimum and maximum range for the subset. If `dimension` is a dimension name, then `range` specifies the range of elements to extract. If `dimension` is a field name, then `range` specifies the range of values to extract. `'Vertical'` subsetting can be used alone or in conjunction with `'Box'` or `'Time'`. To subset a region along multiple dimensions, vertical subsetting can be used up to eight times in one call to `hdfread`.

For example,

hdfread('example.hdf',swath_dataset, 'Fields', fieldname, ...
'Time', {start, stop, 'midpoint'})

Examples

Importing a Data Set by Name

When you know the name of the data set, you can refer to the data set by name in the hdfread command. To read a data set named 'Example SDS', use

data = hdfread('example.hdf', 'Example SDS')

Importing a Data Set Using the Hinfo Structure

When you don't know the name of the data set, follow this procedure.

Use hdfinfo first to retrieve information on the data set.

fileinfo = hdfinfo('example.hdf')
fileinfo = 

    Filename: 'N:\toolbox\matlab\demos\example.hdf'
         SDS: [1x1 struct]
       Vdata: [1x1 struct]

Extract the structure containing information about the particular data set you want to import from fileinfo.

sds_info = fileinfo.SDS
sds_info = 

       Filename: 'N:\toolbox\matlab\demos\example.hdf'
           Type: 'Scientific Data Set'
           Name: 'Example SDS'
           Rank: 2
       DataType: 'int16'
     Attributes: []
           Dims: [2x1 struct]
          Label: {}
    Description: {}
          Index: 0

Pass this structure to hdfread to import the data in the data set.

```
data = hdfread(sds_info)
```

Importing a Subset of a Data Set

You can check the size of the information returned as follows.

sds_info.Dims.Size
ans =
    16
ans =
     5

Using hdfread parameter/value pairs, you can read a subset of the data in the data set. This example specifies a starting index of [3 3], an interval of 1 between values ([] meaning the default value of 1), and a length of 10 rows and 2 columns.

data = hdfread(sds_info, 'Index', {[3 3],[],[10 2]});

data(:,1)
ans =
     7
     8
     9
    10
    11
    12
    13
    14
    15
    16

data(:,2)
ans =
     8
     9
    10
    11
    12
    13
    14
    15
    16
    17

Importing Fields from a Vdata Set

This example retrieves information from example.hdf first, and then reads two fields of the data, Idx and Temp.

info = hdfinfo('example.hdf');

data = hdfread(info.Vdata,...
   'Fields',{'Idx','Temp'})

data = 
    [1x10 int16]
    [1x10 int16]

index = data{1,1};
temp = data{2,1};

temp(1:6)
ans =
     0    12     3     5    10    -1

See Also

hdfinfo, hdf

hdfinfo hdftool