HDF5 Node Interface

• Home
• Documentation
• Install & Setup
• Tutorial
• Writing & Reading Subsets
• Metadata & Attributes
• Events in Packets
• Modify Tables
• Populate Tables
• Working with Images
• Making & Reading Datasets
• Reference
• References
• Library
• Filters
• Dimension Scales
• Globals
• File
• Groups
• Packet Tables
• Datasets
• Tables
• Images

Rank 2 Subsets

This nodejs interface follows the hdf5 example at https://www.hdfgroup.org/HDF5/Tutor/selectsimple.html

If you don’t already have a dataset, start by creating one.

var hdf5 = require('hdf5').hdf5;
var h5lt = require('hdf5').h5lt;

var Access = require('hdf5/lib/globals').Access;

var file = new hdf5.File('./pmc.h5', Access.ACC_TRUNC);
const group=file.createGroup('pmcservices/Double');
const buffer=Buffer.alloc(8*10*8);
for (let j = 0; j < 10; j++) {
    for (let i = 0; i < 8; i++){
            if (j< (10/2)) {
                buffer.writeDoubleLE(1.0, 8*(i*10+j));
            } else {
                buffer.writeDoubleLE(2.0, 8*(i*10+j));
            }
    }
}

h5lt.makeDataset(group.id, 'Waldo', buffer, {type: H5Type.H5T_NATIVE_DOUBLE, rank: 2, rows: 8, columns: 10});

View your h5 file and there should be a dataset with data looking like:

To write a subset use the options on the writeDataset method.

const subsetBuffer=Buffer.alloc(3*4*8, "\0", "binary");
subsetBuffer.type=H5Type.H5T_NATIVE_DOUBLE;
for (let j = 0; j < 4; j++) {
    for (let i = 0; i < 3; i++){
                subsetBuffer.writeDoubleLE(5.0, 8*(i*4+j));
    }
}

h5lt.writeDataset(group.id, 'Waldo', subsetBuffer, {start: [1,2], stride: [1,1], count: [3,4]});

The subset of 5.0’s is down a row and over 2 columns and 4 wide and 3 in height.

To read back a different subset use the readDatasetAsBuffer with the shape defined in the options:

const readAsBuffer=h5lt.readDatasetAsBuffer(group.id, 'Waldo', {start: [3,4], stride: [1,1], count: [2,2]});

The returned nodejs Buffer subset is down 4 row and over 5 columns and 2 wide and 2 in height. The values should be 5.0, 5.0, 1.0 and 2.0.

There is currently only one operator that overwrites the subset data in h5. Remember to close your group to flush the memory to file.