f79207033b
* #8682 Don't log openmp BLAS threads for CUDA Signed-off-by: Alex Black <blacka101@gmail.com> * #8654 Add SameDiff multi-threaded tests Signed-off-by: Alex Black <blacka101@gmail.com> * Switching to op context for SameDiff exec Signed-off-by: Alex Black <blacka101@gmail.com> * Next steps Signed-off-by: Alex Black <blacka101@gmail.com> * Most back to passing Signed-off-by: Alex Black <blacka101@gmail.com> * Fixes Signed-off-by: Alex Black <blacka101@gmail.com> * Better tests, test refactoring Signed-off-by: Alex Black <blacka101@gmail.com> * Small tweak Signed-off-by: Alex Black <blacka101@gmail.com> * Code duplication reduction Signed-off-by: Alex Black <blacka101@gmail.com> * More code deduplication Signed-off-by: Alex Black <blacka101@gmail.com> * CUDA fixes Signed-off-by: Alex Black <blacka101@gmail.com> * More CUDA fixes Signed-off-by: Alex Black <blacka101@gmail.com> * More fixes Signed-off-by: Alex Black <blacka101@gmail.com> * Small fix Signed-off-by: Alex Black <blacka101@gmail.com> * ND4S small fixes Signed-off-by: Alex Black <blacka101@gmail.com> |
||
---|---|---|
.. | ||
project | ||
src | ||
.gitignore | ||
.scalafmt.conf | ||
.travis.yml | ||
LICENSE | ||
README.md | ||
build.sbt | ||
pom.xml | ||
sbt-pom.xml |
README.md
ND4S: Scala bindings for ND4J
ND4S is open-source Scala bindings for ND4J. Released under an Apache 2.0 license.
Main Features
- NDArray manipulation syntax sugar with safer type.
- NDArray slicing syntax, similar with NumPy.
Installation
Install via Maven
ND4S is already included in official Maven repositories.
With IntelliJ, incorporation of ND4S is easy: just create a new Scala project, go to "Project Settings"/Libraries, add "From Maven...", and search for nd4s.
As an alternative, one may simply add the line below to build.sbt
and re-build project.
val nd4jVersion = "1.0.0-alpha"
libraryDependencies += "org.nd4j" % "nd4j-native-platform" % nd4jVersion
libraryDependencies += "org.nd4j" %% "nd4s" % nd4jVersion
One may want to check our maven repository page and replace 1.0.0-alpha
with the latest version.
No need for git-cloning & compiling!
Clone from the GitHub Repo
ND4S is actively developed. You can clone the repository, compile it, and reference it in your project.
Clone the repository:
$ git clone https://github.com/eclipse/deeplearning4j.git
Compile the project:
$ cd nd4s
$ sbt +publish-local
Try ND4S in REPL
The easiest way to play ND4S around is cloning this repository and run the following command.
$ cd nd4s
$ sbt test:console
It starts REPL with importing org.nd4s.Implicits._
and org.nd4j.linalg.factory.Nd4j
automatically. It uses jblas backend at default.
scala> val arr = (1 to 9).asNDArray(3,3)
arr: org.nd4j.linalg.api.ndarray.INDArray =
[[1.00,2.00,3.00]
[4.00,5.00,6.00]
[7.00,8.00,9.00]]
scala> val sub = arr(0->2,1->3)
sub: org.nd4j.linalg.api.ndarray.INDArray =
[[2.00,3.00]
[5.00,6.00]]
CheatSheet(WIP)
ND4S syntax | Equivalent NumPy syntax | Result |
---|---|---|
Array(Array(1,2,3),Array(4,5,6)).toNDArray | np.array(1, 2 , 3], [4, 5, 6) | 1.0, 2.0, 3.0] [4.0, 5.0, 6.0 |
val arr = (1 to 9).asNDArray(3,3) | arr = np.arange(1,10).reshape(3,3) | 1.0, 2.0, 3.0] [4.0, 5.0, 6.0] ,[7.0, 8.0, 9.0 |
arr(0,0) | arr[0,0] | 1.0 |
arr(0,->) | arr[0,:] | [1.0, 2.0, 3.0] |
arr(--->) | arr[...] | 1.0, 2.0, 3.0] [4.0, 5.0, 6.0] ,[7.0, 8.0, 9.0 |
arr(0 -> 3 by 2, ->) | arr[0:3:2,:] | 1.0, 2.0, 3.0] [7.0, 8.0, 9.0 |
arr(0 to 2 by 2, ->) | arr[0:3:2,:] | 1.0, 2.0, 3.0] [7.0, 8.0, 9.0 |
arr.filter(_ > 3) | np.where(arr > 3, arr, 0) | 0.0, 0.0, 0.0] [4.0, 5.0, 6.0] ,[7.0, 8.0, 9.0 |
arr.map(_ % 3) | 1.0, 2.0, 0.0] [1.0, 2.0, 0.0] ,[1.0, 2.0, 0.0 | |
arr.filterBit(_ < 4) | 1.0, 1.0, 1.0] [0.0, 0.0, 0.0] ,[0.0, 0.0, 0.0 | |
arr + arr | arr + arr | 2.0, 4.0, 6.0] [8.0, 10.0, 12.0] ,[14.0, 16.0, 18.0 |
arr * arr | arr * arr | 1.0, 4.0, 9.0] [16.0, 25.0, 36.0] ,[49.0, 64.0, 81.0 |
arr dot arr | np.dot(arr, arr) | 30.0, 36.0, 42.0] [66.0, 81.0, 96.0] ,[102.0, 126.0, 150.0 |
arr.sumT | np.sum(arr) | 45.0 //returns Double value |
val comp = Array(1 + i, 1 + 2 * i).toNDArray | comp = np.array([1 + 1j, 1 + 2j]) | [1.0 + 1.0i ,1.0 + 2.0i] |
comp.sumT | np.sum(comp) | 2.0 + 3.0i //returns IComplexNumber value |
for(row <- arr.rowP if row.get(0) > 1) yield row*2 | 8.00,10.00,12.00] [14.00,16.00,18.00 | |
val tensor = (1 to 8).asNDArray(2,2,2) | tensor = np.arange(1,9).reshape(2,2,2) | [1.00,2.00] [3.00,4.00 5.00,6.00] [7.00,8.00] |
for(slice <- tensor.sliceP if slice.get(0) > 1) yield slice*2 | [10.00,12.00][14.00,16.00] | |
arr(0 -> 3 by 2, ->) = 0 | 0.00,0.00,0.00] [4.00,5.00,6.00] [0.00,0.00,0.00 |