public class MxSparseNDArray extends MxNDArray
MxSparseNDArray is an instance of MxNDArray and NDArray for sparse
NDArrays.
MxSparseNDArrays are created automatically when the Engine creates Arrays that are
sparse. They can be created deliberately by specifying the SparseFormat. Some operations
may not be supported with Sparse NDArrays in MXNet.
SparseFormathandle| Modifier and Type | Method and Description |
|---|---|
ai.djl.ndarray.NDArray |
get(ai.djl.ndarray.index.NDIndex index) |
void |
set(java.nio.Buffer data) |
java.nio.ByteBuffer |
toByteBuffer() |
abs, acos, acosh, add, add, addi, addi, argMax, argMax, argMin, argMin, argSort, asin, asinh, atan, atanh, attach, attachGradient, attachGradient, backward, booleanMask, broadcast, cbrt, ceil, clip, close, contentEquals, contentEquals, copyTo, cos, cosh, cumSum, cumSum, detach, div, div, divi, divi, dot, eq, eq, equals, exp, expandDims, flatten, floor, getDataType, getDevice, getGradient, getManager, getName, getNDArrayInternal, getShape, getSparseFormat, gt, gt, gte, gte, hasGradient, hashCode, isInfinite, isNaN, log, log10, log2, logicalAnd, logicalNot, logicalOr, logicalXor, logSoftmax, lt, lt, lte, lte, matMul, max, max, max, maximum, maximum, mean, mean, median, median, min, min, minimum, minimum, mod, mod, modi, modi, mul, mul, muli, muli, neg, negi, neq, neq, nonzero, onesLike, percentile, percentile, pow, pow, powi, powi, prod, prod, repeat, repeat, repeat, repeat, reshape, round, sequenceMask, sequenceMask, setName, setShouldFree, sin, sinh, softmax, sort, sort, split, sqrt, square, squeeze, squeeze, sub, sub, subi, subi, sum, sum, swapAxes, tan, tanh, tile, tile, tile, tile, toDegrees, toDense, toDevice, toRadians, toSparse, toString, toType, trace, transpose, transpose, trunc, waitAll, waitToRead, waitToWrite, zerosLikefinalize, getHandle, getUid, isReleasedall, allClose, allClose, any, argSort, argSort, booleanMask, broadcast, concat, concat, countNonzero, countNonzero, decode, duplicate, encode, get, get, get, getBoolean, getByte, getDouble, getFloat, getInt, getLong, getScalar, getUid, getUint8, isEmpty, isScalar, isSparse, like, mean, min, none, prod, reshape, set, set, set, set, set, set, set, set, setScalar, shapeEquals, size, size, split, split, split, squeeze, stack, stack, sum, toArray, toBooleanArray, toByteArray, toDebugString, toDoubleArray, toFloatArray, toIntArray, toLongArray, toUint8Array, trace, tracepublic void set(java.nio.Buffer data)
public ai.djl.ndarray.NDArray get(ai.djl.ndarray.index.NDIndex index)
public java.nio.ByteBuffer toByteBuffer()
toByteBuffer in interface ai.djl.ndarray.NDArraytoByteBuffer in class MxNDArray