Skip to content
/ npy Public

npy is a Java 11+ library for reading and writing arrays in the NumPy NPY and NPZ format.

License

Notifications You must be signed in to change notification settings

GreenDelta/npy

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

npy

npy is a Java 11+ library for reading and writing arrays in the NumPy NPY and NPZ format.

Features

  • written in pure Java 11 with no other dependencies
  • supports reading arrays of 8-, 16-, 32-, and 64-bit signed and unsigned integers as well as booleans and 16-, 32-, and 64-bit floating point numbers
  • supports writing arrays of all primitive Java types: boolean, byte, short, int, long, float, double, char
  • supports all NPY format versions (1, 2, and 3)
  • provides a simple NpyArray interface for all array types and conversions as well as a utility class Array2d for reading matrix data

Limitations

  • only supports types listed above, no tuples, structs, etc.
  • needs more tests, version is 0.0.1

Alternatives

Installation

The latest version is available on the Maven Central. If you are using Maven add the following dependency to your pom.xml:

<dependency>
  <groupId>org.openlca</groupId>
  <artifactId>npy</artifactId>
  <version>0.0.1</version>
</dependency>

For Gradle, add this to your build.gradle:

dependencies {
  compile 'org.openlca:npy:0.0.1'
}

Examples

Reading and writing an NPY file:

var array = Npy.read(file);
Npy.write(otherFile, array);

Type checks and type conversions:

if (array.isDoubleArray()) {
  var doubles = array.asDoubleArray();
  for (double value : doubles.values()) {
    // ...    
  }
  var ints = doubles.asIntArray();  // may result in data loss
  for (int values : ints.values()) {
    // ...
  }
  var longs = ints.asLongArray();
  // ...
}

Reading and writing NPZ files:

for (var entry : Npz.entries(npzFile)) {
  var array = Npz.load(npzFile, entry);
  // ...
}
Npz.create(npzFile, out -> {
  Npz.write(out, "shape.npy", shape);	
  Npz.write(out, "values.npy", values);
  // ...
});

Reading parts of a file; e.g. the first 100 elements:

var range = Npy.readRange(file, numberOfElements, offset);

Using the Array2d utilities for 2-dimensional arrays (matrices):

Npy.use(file, (chan, head) -> {
  var row42 = Array2d.readRow(chan, head, 142);
  var col42 = Array2d.readColumn(chan, head, 42);
  // ...
});

About

npy is a Java 11+ library for reading and writing arrays in the NumPy NPY and NPZ format.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published