INSTRUCTION
stringlengths
1
46.3k
RESPONSE
stringlengths
75
80.2k
Returns a dispatch code for calling basic or advanced indexing functions.
def _get_indexing_dispatch_code(key): """Returns a dispatch code for calling basic or advanced indexing functions.""" if isinstance(key, (NDArray, np.ndarray)): return _NDARRAY_ADVANCED_INDEXING elif isinstance(key, list): # TODO(junwu): Add support for nested lists besides integer list ...
Given start, stop, step and array length, return absolute values of start, stop, and step for generating index range. The returned values have been compensated by adding length if they are less than zero for all the cases but slice(None, None, -1). Note that the returned value of stop is not necessarily...
def _get_index_range(start, stop, length, step=1): """Given start, stop, step and array length, return absolute values of start, stop, and step for generating index range. The returned values have been compensated by adding length if they are less than zero for all the cases but slice(None, None, -1). ...
Given data and index shapes, get the output `NDArray` shape. This basically implements the infer shape logic of op gather_nd.
def _get_oshape_of_gather_nd_op(dshape, ishape): """Given data and index shapes, get the output `NDArray` shape. This basically implements the infer shape logic of op gather_nd.""" assert len(dshape) > 0 and len(ishape) > 0 oshape = list(ishape[1:]) if ishape[0] < len(dshape): oshape.extend(...
Given start, stop, and stop, calculate the number of elements of this slice.
def _get_dim_size(start, stop, step): """Given start, stop, and stop, calculate the number of elements of this slice.""" assert step != 0 if step > 0: assert start < stop dim_size = (stop - start - 1) // step + 1 else: assert stop < start dim_size = (start - stop - 1)...
Given two shapes that are not identical, find the shape that both input shapes can broadcast to.
def _get_broadcast_shape(shape1, shape2): """Given two shapes that are not identical, find the shape that both input shapes can broadcast to.""" if shape1 == shape2: return shape1 length1 = len(shape1) length2 = len(shape2) if length1 > length2: shape = list(shape1) else: ...
Returns a new array filled with all ones, with the given shape and type. Parameters ---------- shape : int or tuple of int or list of int The shape of the empty array. ctx : Context, optional An optional device context. Defaults to the current default context (``mxnet.context.cu...
def ones(shape, ctx=None, dtype=None, **kwargs): """Returns a new array filled with all ones, with the given shape and type. Parameters ---------- shape : int or tuple of int or list of int The shape of the empty array. ctx : Context, optional An optional device context. Def...
Returns a new array of given shape and type, filled with the given value `val`. Parameters -------- shape : int or tuple of int The shape of the new array. val : scalar Fill value. ctx : Context, optional Device context (default is the current default context). dtype : `...
def full(shape, val, ctx=None, dtype=mx_real_t, out=None): """Returns a new array of given shape and type, filled with the given value `val`. Parameters -------- shape : int or tuple of int The shape of the new array. val : scalar Fill value. ctx : Context, optional Devi...
Creates an array from any object exposing the array interface. Parameters ---------- source_array : array_like An object exposing the array interface, an object whose `__array__` method returns an array, or any (nested) sequence. ctx : Context, optional Device context (default i...
def array(source_array, ctx=None, dtype=None): """Creates an array from any object exposing the array interface. Parameters ---------- source_array : array_like An object exposing the array interface, an object whose `__array__` method returns an array, or any (nested) sequence. ctx...
Moves the `source` axis into the `destination` position while leaving the other axes in their original order Parameters ---------- tensor : mx.nd.array The array which axes should be reordered source : int or sequence of int Original position of the axes to move. Can be negative but...
def moveaxis(tensor, source, destination): """Moves the `source` axis into the `destination` position while leaving the other axes in their original order Parameters ---------- tensor : mx.nd.array The array which axes should be reordered source : int or sequence of int Original...
Returns evenly spaced values within a given interval. Values are generated within the half-open interval [`start`, `stop`). In other words, the interval includes `start` but excludes `stop`. The function is similar to the built-in Python function `range` and to `numpy.arange`, but returns an `NDArray`....
def arange(start, stop=None, step=1.0, repeat=1, infer_range=None, ctx=None, dtype=mx_real_t): """Returns evenly spaced values within a given interval. Values are generated within the half-open interval [`start`, `stop`). In other words, the interval includes `start` but excludes `stop`. The function is ...
Helper function for element-wise operation. The function will perform numpy-like broadcasting if needed and call different functions. Parameters -------- lhs : NDArray or numeric value Left-hand side operand. rhs : NDArray or numeric value Right-hand operand, fn_array : functi...
def _ufunc_helper(lhs, rhs, fn_array, fn_scalar, lfn_scalar, rfn_scalar=None): """ Helper function for element-wise operation. The function will perform numpy-like broadcasting if needed and call different functions. Parameters -------- lhs : NDArray or numeric value Left-hand side operand....
Returns element-wise sum of the input arrays with broadcasting. Equivalent to ``lhs + rhs``, ``mx.nd.broadcast_add(lhs, rhs)`` and ``mx.nd.broadcast_plus(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the arrays are br...
def add(lhs, rhs): """Returns element-wise sum of the input arrays with broadcasting. Equivalent to ``lhs + rhs``, ``mx.nd.broadcast_add(lhs, rhs)`` and ``mx.nd.broadcast_plus(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, ...
Returns element-wise difference of the input arrays with broadcasting. Equivalent to ``lhs - rhs``, ``mx.nd.broadcast_sub(lhs, rhs)`` and ``mx.nd.broadcast_minus(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the array...
def subtract(lhs, rhs): """Returns element-wise difference of the input arrays with broadcasting. Equivalent to ``lhs - rhs``, ``mx.nd.broadcast_sub(lhs, rhs)`` and ``mx.nd.broadcast_minus(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them ha...
Returns element-wise product of the input arrays with broadcasting. Equivalent to ``lhs * rhs`` and ``mx.nd.broadcast_mul(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the arrays are broadcastable to a common shape. ...
def multiply(lhs, rhs): """Returns element-wise product of the input arrays with broadcasting. Equivalent to ``lhs * rhs`` and ``mx.nd.broadcast_mul(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the arrays are broadca...
Returns element-wise division of the input arrays with broadcasting. Equivalent to ``lhs / rhs`` and ``mx.nd.broadcast_div(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the arrays are broadcastable to a common shape. ...
def divide(lhs, rhs): """Returns element-wise division of the input arrays with broadcasting. Equivalent to ``lhs / rhs`` and ``mx.nd.broadcast_div(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the arrays are broadcas...
Returns element-wise modulo of the input arrays with broadcasting. Equivalent to ``lhs % rhs`` and ``mx.nd.broadcast_mod(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the arrays are broadcastable to a common shape. P...
def modulo(lhs, rhs): """Returns element-wise modulo of the input arrays with broadcasting. Equivalent to ``lhs % rhs`` and ``mx.nd.broadcast_mod(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the arrays are broadcasta...
Returns result of first array elements raised to powers from second array, element-wise with broadcasting. Equivalent to ``base ** exp`` and ``mx.nd.broadcast_power(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the ar...
def power(base, exp): """Returns result of first array elements raised to powers from second array, element-wise with broadcasting. Equivalent to ``base ** exp`` and ``mx.nd.broadcast_power(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them h...
Returns element-wise maximum of the input arrays with broadcasting. Equivalent to ``mx.nd.broadcast_maximum(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the arrays are broadcastable to a common shape. Parameters ...
def maximum(lhs, rhs): """Returns element-wise maximum of the input arrays with broadcasting. Equivalent to ``mx.nd.broadcast_maximum(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the arrays are broadcastable to a com...
Returns element-wise minimum of the input arrays with broadcasting. Equivalent to ``mx.nd.broadcast_minimum(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the arrays are broadcastable to a common shape. Parameters ...
def minimum(lhs, rhs): """Returns element-wise minimum of the input arrays with broadcasting. Equivalent to ``mx.nd.broadcast_minimum(lhs, rhs)``. .. note:: If the corresponding dimensions of two arrays have the same size or one of them has size 1, then the arrays are broadcastable to a com...
Returns the result of element-wise **not equal to** (!=) comparison operation with broadcasting. For each element in input arrays, return 1(true) if corresponding elements are different, otherwise return 0(false). Equivalent to ``lhs != rhs`` and ``mx.nd.broadcast_not_equal(lhs, rhs)``. .. note::...
def not_equal(lhs, rhs): """Returns the result of element-wise **not equal to** (!=) comparison operation with broadcasting. For each element in input arrays, return 1(true) if corresponding elements are different, otherwise return 0(false). Equivalent to ``lhs != rhs`` and ``mx.nd.broadcast_not_e...
Returns the result of element-wise **greater than** (>) comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements are greater than rhs, otherwise return 0(false). Equivalent to ``lhs > rhs`` and ``mx.nd.broadcast_greater(lhs, rhs)``. .. note:: ...
def greater(lhs, rhs): """Returns the result of element-wise **greater than** (>) comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements are greater than rhs, otherwise return 0(false). Equivalent to ``lhs > rhs`` and ``mx.nd.broadcast_greater(lhs,...
Returns the result of element-wise **greater than or equal to** (>=) comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements are greater than equal to rhs, otherwise return 0(false). Equivalent to ``lhs >= rhs`` and ``mx.nd.broadcast_greater_equal(lhs, ...
def greater_equal(lhs, rhs): """Returns the result of element-wise **greater than or equal to** (>=) comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements are greater than equal to rhs, otherwise return 0(false). Equivalent to ``lhs >= rhs`` and `...
Returns the result of element-wise **lesser than** (<) comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements are less than rhs, otherwise return 0(false). Equivalent to ``lhs < rhs`` and ``mx.nd.broadcast_lesser(lhs, rhs)``. .. note:: If ...
def lesser(lhs, rhs): """Returns the result of element-wise **lesser than** (<) comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements are less than rhs, otherwise return 0(false). Equivalent to ``lhs < rhs`` and ``mx.nd.broadcast_lesser(lhs, rhs)`...
Returns the result of element-wise **lesser than or equal to** (<=) comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements are lesser than equal to rhs, otherwise return 0(false). Equivalent to ``lhs <= rhs`` and ``mx.nd.broadcast_lesser_equal(lhs, rhs...
def lesser_equal(lhs, rhs): """Returns the result of element-wise **lesser than or equal to** (<=) comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements are lesser than equal to rhs, otherwise return 0(false). Equivalent to ``lhs <= rhs`` and ``mx...
Returns the result of element-wise **logical and** comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements and rhs elements are true, otherwise return 0(false). Equivalent to ``lhs and rhs`` and ``mx.nd.broadcast_logical_and(lhs, rhs)``. .. note:: ...
def logical_and(lhs, rhs): """Returns the result of element-wise **logical and** comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements and rhs elements are true, otherwise return 0(false). Equivalent to ``lhs and rhs`` and ``mx.nd.broadcast_logica...
Returns the result of element-wise **logical or** comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements or rhs elements are true, otherwise return 0(false). Equivalent to ``lhs or rhs`` and ``mx.nd.broadcast_logical_or(lhs, rhs)``. .. note:: ...
def logical_or(lhs, rhs): """Returns the result of element-wise **logical or** comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements or rhs elements are true, otherwise return 0(false). Equivalent to ``lhs or rhs`` and ``mx.nd.broadcast_logical_or...
Returns the result of element-wise **logical xor** comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements or rhs elements are true, otherwise return 0(false). Equivalent to ``bool(lhs) ^ bool(rhs)`` and ``mx.nd.broadcast_logical_xor(lhs, rhs)``. ....
def logical_xor(lhs, rhs): """Returns the result of element-wise **logical xor** comparison operation with broadcasting. For each element in input arrays, return 1(true) if lhs elements or rhs elements are true, otherwise return 0(false). Equivalent to ``bool(lhs) ^ bool(rhs)`` and ``mx.nd.broadca...
DEPRECATED, use ``concat`` instead Parameters ---------- arrays : list of `NDArray` Arrays to be concatenate. They must have identical shape except the first dimension. They also must have the same data type. axis : int The axis along which to concatenate. always_copy : bool...
def concatenate(arrays, axis=0, always_copy=True): """DEPRECATED, use ``concat`` instead Parameters ---------- arrays : list of `NDArray` Arrays to be concatenate. They must have identical shape except the first dimension. They also must have the same data type. axis : int T...
DEPRECATED, use mx.img instead Parameters ---------- str_img : str Binary image data clip_rect : iterable of 4 int Clip decoded image to rectangle (x0, y0, x1, y1). out : NDArray Output buffer. Can be 3 dimensional (c, h, w) or 4 dimensional (n, c, h, w). index : int ...
def imdecode(str_img, clip_rect=(0, 0, 0, 0), out=None, index=0, channels=3, mean=None): """DEPRECATED, use mx.img instead Parameters ---------- str_img : str Binary image data clip_rect : iterable of 4 int Clip decoded image to rectangle (x0, y0, x1, y1). out : NDArray ...
Returns a new array filled with all zeros, with the given shape and type. Parameters ---------- shape : int or tuple of int The shape of the empty array. ctx : Context, optional An optional device context (default is the current default context). dtype : str or numpy.dtype, optional...
def zeros(shape, ctx=None, dtype=None, **kwargs): """Returns a new array filled with all zeros, with the given shape and type. Parameters ---------- shape : int or tuple of int The shape of the empty array. ctx : Context, optional An optional device context (default is the current d...
Return a 2-D array with ones on the diagonal and zeros elsewhere. Parameters ---------- N: int Number of rows in the output. M: int, optional Number of columns in the output. If 0, defaults to N. k: int, optional Index of the diagonal: 0 (the default) refers to the main diag...
def eye(N, M=0, k=0, ctx=None, dtype=None, **kwargs): """Return a 2-D array with ones on the diagonal and zeros elsewhere. Parameters ---------- N: int Number of rows in the output. M: int, optional Number of columns in the output. If 0, defaults to N. k: int, optional I...
Returns a new array of given shape and type, without initializing entries. Parameters ---------- shape : int or tuple of int The shape of the empty array. ctx : Context, optional An optional device context (default is the current default context). dtype : str or numpy.dtype, optiona...
def empty(shape, ctx=None, dtype=None): """Returns a new array of given shape and type, without initializing entries. Parameters ---------- shape : int or tuple of int The shape of the empty array. ctx : Context, optional An optional device context (default is the current default co...
Compute the histogram of the input data. Parameters ---------- a : NDArray Input data. The histogram is computed over the flattened array. bins : int or sequence of scalars If bins is an int, it defines the number of equal-width bins in the given range (10, by default). If bins ...
def histogram(a, bins=10, range=None): """Compute the histogram of the input data. Parameters ---------- a : NDArray Input data. The histogram is computed over the flattened array. bins : int or sequence of scalars If bins is an int, it defines the number of equal-width bins in the ...
Split an array into multiple sub-arrays. Parameters ---------- ary : NDArray Array to be divided into sub-arrays. indices_or_sections : int or tuple of ints If `indices_or_sections` is an integer, N, the array will be divided into N equal arrays along `axis`. If such a split is...
def split_v2(ary, indices_or_sections, axis=0, squeeze_axis=False): """Split an array into multiple sub-arrays. Parameters ---------- ary : NDArray Array to be divided into sub-arrays. indices_or_sections : int or tuple of ints If `indices_or_sections` is an integer, N, the array wi...
Returns a reference view of NDArray that represents as DLManagedTensor until all previous write operations on the current array are finished. Parameters ---------- data: NDArray input data. Returns ------- PyCapsule (the pointer of DLManagedTensor) a reference view of ND...
def to_dlpack_for_read(data): """Returns a reference view of NDArray that represents as DLManagedTensor until all previous write operations on the current array are finished. Parameters ---------- data: NDArray input data. Returns ------- PyCapsule (the pointer of DLManagedT...
Returns a reference view of NDArray that represents as DLManagedTensor until all previous read/write operations on the current array are finished. Parameters ---------- data: NDArray input data. Returns ------- PyCapsule (the pointer of DLManagedTensor) a reference view ...
def to_dlpack_for_write(data): """Returns a reference view of NDArray that represents as DLManagedTensor until all previous read/write operations on the current array are finished. Parameters ---------- data: NDArray input data. Returns ------- PyCapsule (the pointer of DLMa...
Returns a NDArray backed by a dlpack tensor. Parameters ---------- dlpack: PyCapsule (the pointer of DLManagedTensor) input data Returns ------- NDArray a NDArray backed by a dlpack tensor Examples -------- >>> x = mx.nd.ones((2,3)) >>> y = mx.nd.to_dlpack_for_...
def from_dlpack(dlpack): """Returns a NDArray backed by a dlpack tensor. Parameters ---------- dlpack: PyCapsule (the pointer of DLManagedTensor) input data Returns ------- NDArray a NDArray backed by a dlpack tensor Examples -------- >>> x = mx.nd.ones((2,3)) ...
Returns an MXNet's NDArray backed by Numpy's ndarray. Parameters ---------- ndarray: numpy.ndarray input data zero_copy: bool Whether we use DLPack's zero-copy conversion to convert to MXNet's NDArray. This is only available for c-contiguous arrays, i.e. array.flags[C_CONTIGUOU...
def from_numpy(ndarray, zero_copy=True): """Returns an MXNet's NDArray backed by Numpy's ndarray. Parameters ---------- ndarray: numpy.ndarray input data zero_copy: bool Whether we use DLPack's zero-copy conversion to convert to MXNet's NDArray. This is only available for c...
Returns an index array for use in scatter_nd and gather_nd.
def _get_index_nd(self, key): """Returns an index array for use in scatter_nd and gather_nd.""" def _is_advanced_index(index): """The definition of advanced index here includes integers as well, while integers are considered as basic index type when the key contains only ...
Given value and vshape, create an `NDArray` from value with the same context and dtype as the current one and broadcast it to vshape.
def _prepare_value_nd(self, value, vshape): """Given value and vshape, create an `NDArray` from value with the same context and dtype as the current one and broadcast it to vshape.""" if isinstance(value, numeric_types): value_nd = full(shape=vshape, val=value, ctx=self.context, dtyp...
This function is called by __setitem__ when key is an advanced index.
def _set_nd_advanced_indexing(self, key, value): """This function is called by __setitem__ when key is an advanced index.""" indices = self._get_index_nd(key) vshape = _get_oshape_of_gather_nd_op(self.shape, indices.shape) value_nd = self._prepare_value_nd(value, vshape) _interna...
This function is called by __setitem__ when key is a basic index, i.e. an integer, or a slice, or a tuple of integers and slices. No restrictions on the values of slices' steps.
def _set_nd_basic_indexing(self, key, value): """This function is called by __setitem__ when key is a basic index, i.e. an integer, or a slice, or a tuple of integers and slices. No restrictions on the values of slices' steps.""" shape = self.shape if isinstance(key, integer_type...
This function is called when key is a slice, or an integer, or a tuple of slices or integers
def _get_nd_basic_indexing(self, key): """This function is called when key is a slice, or an integer, or a tuple of slices or integers""" shape = self.shape if isinstance(key, integer_types): if key > shape[0] - 1: raise IndexError( 'index ...
Performs a synchronized copy from the `source_array` to the current array. This is called through ``x[:] = source_array``, where the `source_array` is a `numpy.ndarray` or array-like object. This function blocks until all the pending read/write operations with respect to the current `NDA...
def _sync_copyfrom(self, source_array): """Performs a synchronized copy from the `source_array` to the current array. This is called through ``x[:] = source_array``, where the `source_array` is a `numpy.ndarray` or array-like object. This function blocks until all the pending read/write ...
Returns a sliced NDArray that shares memory with the current one. This is called through ``x[start:stop]``. Parameters ---------- start : int Starting inclusive index of slice in the first dim. stop : int Finishing exclusive index of slice in the first di...
def _slice(self, start, stop): """Returns a sliced NDArray that shares memory with the current one. This is called through ``x[start:stop]``. Parameters ---------- start : int Starting inclusive index of slice in the first dim. stop : int Finishin...
Returns a view of the array sliced at `idx` in the first dim. This is called through ``x[idx]``. Parameters ---------- idx : int index for slicing the `NDArray` in the first dim. Returns ------- NDArray `NDArray` sharing the memory with t...
def _at(self, idx): """Returns a view of the array sliced at `idx` in the first dim. This is called through ``x[idx]``. Parameters ---------- idx : int index for slicing the `NDArray` in the first dim. Returns ------- NDArray `NDA...
Returns a **view** of this array with a new shape without altering any data. Parameters ---------- shape : tuple of int, or n ints The new shape should not change the array size, namely ``np.prod(new_shape)`` should be equal to ``np.prod(self.shape)``. Some d...
def reshape(self, *shape, **kwargs): """Returns a **view** of this array with a new shape without altering any data. Parameters ---------- shape : tuple of int, or n ints The new shape should not change the array size, namely ``np.prod(new_shape)`` should be equa...
Broadcasts the input array to a new shape. Broadcasting is only allowed on axes with size 1. The new shape cannot change the number of dimensions. For example, you could broadcast from shape (2, 1) to (2, 3), but not from shape (2, 3) to (2, 3, 3). Parameters ----------...
def broadcast_to(self, shape): """Broadcasts the input array to a new shape. Broadcasting is only allowed on axes with size 1. The new shape cannot change the number of dimensions. For example, you could broadcast from shape (2, 1) to (2, 3), but not from shape (2, 3) to (2, 3, ...
Tuple of array dimensions. Examples -------- >>> x = mx.nd.array([1, 2, 3, 4]) >>> x.shape (4L,) >>> y = mx.nd.zeros((2, 3, 4)) >>> y.shape (2L, 3L, 4L)
def shape(self): """Tuple of array dimensions. Examples -------- >>> x = mx.nd.array([1, 2, 3, 4]) >>> x.shape (4L,) >>> y = mx.nd.zeros((2, 3, 4)) >>> y.shape (2L, 3L, 4L) """ ndim = mx_int() pdata = ctypes.POINTER(mx_int)...
Device context of the array. Examples -------- >>> x = mx.nd.array([1, 2, 3, 4]) >>> x.context cpu(0) >>> type(x.context) <class 'mxnet.context.Context'> >>> y = mx.nd.zeros((2,3), mx.gpu(0)) >>> y.context gpu(0)
def context(self): """Device context of the array. Examples -------- >>> x = mx.nd.array([1, 2, 3, 4]) >>> x.context cpu(0) >>> type(x.context) <class 'mxnet.context.Context'> >>> y = mx.nd.zeros((2,3), mx.gpu(0)) >>> y.context gpu...
Data-type of the array's elements. Returns ------- numpy.dtype This NDArray's data type. Examples -------- >>> x = mx.nd.zeros((2,3)) >>> x.dtype <type 'numpy.float32'> >>> y = mx.nd.zeros((2,3), dtype='int32') >>> y.dtype ...
def dtype(self): """Data-type of the array's elements. Returns ------- numpy.dtype This NDArray's data type. Examples -------- >>> x = mx.nd.zeros((2,3)) >>> x.dtype <type 'numpy.float32'> >>> y = mx.nd.zeros((2,3), dtype='int...
Whether this array's corresponding gradient array (registered via `autograd.mark_variables`) has been updated by `autograd.backward` since last reset. `_fresh_grad` need to be manually set to False after consuming gradient (usually after updating this array).
def _fresh_grad(self): """Whether this array's corresponding gradient array (registered via `autograd.mark_variables`) has been updated by `autograd.backward` since last reset. `_fresh_grad` need to be manually set to False after consuming gradient (usually after updating this ...
Returns a ``numpy.ndarray`` object with value copied from this array. Examples -------- >>> x = mx.nd.ones((2,3)) >>> y = x.asnumpy() >>> type(y) <type 'numpy.ndarray'> >>> y array([[ 1., 1., 1.], [ 1., 1., 1.]], dtype=float32) ...
def asnumpy(self): """Returns a ``numpy.ndarray`` object with value copied from this array. Examples -------- >>> x = mx.nd.ones((2,3)) >>> y = x.asnumpy() >>> type(y) <type 'numpy.ndarray'> >>> y array([[ 1., 1., 1.], [ 1., 1., ...
Returns a copy of the array after casting to a specified type. Parameters ---------- dtype : numpy.dtype or str The type of the returned array. copy : bool Default `True`. By default, astype always returns a newly allocated ndarray on the same context...
def astype(self, dtype, copy=True): """Returns a copy of the array after casting to a specified type. Parameters ---------- dtype : numpy.dtype or str The type of the returned array. copy : bool Default `True`. By default, astype always returns a newly ...
Copies the value of this array to another array. If ``other`` is a ``NDArray`` object, then ``other.shape`` and ``self.shape`` should be the same. This function copies the value from ``self`` to ``other``. If ``other`` is a context, a new ``NDArray`` will be first created on th...
def copyto(self, other): """Copies the value of this array to another array. If ``other`` is a ``NDArray`` object, then ``other.shape`` and ``self.shape`` should be the same. This function copies the value from ``self`` to ``other``. If ``other`` is a context, a new ``NDArray``...
Returns an array on the target device with the same value as this array. If the target context is the same as ``self.context``, then ``self`` is returned. Otherwise, a copy is made. Parameters ---------- context : Context The target context. Returns ...
def as_in_context(self, context): """Returns an array on the target device with the same value as this array. If the target context is the same as ``self.context``, then ``self`` is returned. Otherwise, a copy is made. Parameters ---------- context : Context ...
Attach a gradient buffer to this NDArray, so that `backward` can compute gradient with respect to it. Parameters ---------- grad_req : {'write', 'add', 'null'} How gradient will be accumulated. - 'write': gradient will be overwritten on every backward. ...
def attach_grad(self, grad_req='write', stype=None): """Attach a gradient buffer to this NDArray, so that `backward` can compute gradient with respect to it. Parameters ---------- grad_req : {'write', 'add', 'null'} How gradient will be accumulated. - 'wr...
Returns gradient buffer attached to this NDArray.
def grad(self): """Returns gradient buffer attached to this NDArray.""" from . import _ndarray_cls hdl = NDArrayHandle() check_call(_LIB.MXNDArrayGetGrad(self.handle, ctypes.byref(hdl))) if hdl.value is None: return None return _ndarray_cls(hdl)
Returns a new NDArray, detached from the current graph.
def detach(self): """Returns a new NDArray, detached from the current graph.""" from . import _ndarray_cls hdl = NDArrayHandle() check_call(_LIB.MXNDArrayDetach(self.handle, ctypes.byref(hdl))) return _ndarray_cls(hdl)
Compute the gradients of this NDArray w.r.t variables. Parameters ---------- out_grad : NDArray, optional Gradient with respect to head. retain_graph : bool, optional Whether to retain the computaion graph for another backward pass on the same graph. ...
def backward(self, out_grad=None, retain_graph=False, train_mode=True): """Compute the gradients of this NDArray w.r.t variables. Parameters ---------- out_grad : NDArray, optional Gradient with respect to head. retain_graph : bool, optional Whether to re...
Build the align array
def build(self, align_path): """ Build the align array """ file = open(align_path, 'r') lines = file.readlines() file.close() # words: list([op, ed, word]) words = [] for line in lines: _op, _ed, word = line.strip().split(' ') ...
Get sentence
def sentence(self, padding=75): """ Get sentence """ vec = word_to_vector(self.sentence_str) vec += [-1] * (padding - self.sentence_length) return np.array(vec, dtype=np.int32)
Get words
def word(self, _id, padding=75): """ Get words """ word = self.words[_id][2] vec = word_to_vector(word) vec += [-1] * (padding - len(vec)) return np.array(vec, dtype=np.int32)
Get the position of words
def word_frame_pos(self, _id): """ Get the position of words """ left = int(self.words[_id][0]/1000) right = max(left+1, int(self.words[_id][1]/1000)) return (left, right)
Prepares the module for processing a data batch by pulling row_sparse parameters from kvstore to all devices based on rowids. Parameters ---------- param_rowids : dict of str to NDArray of list of NDArrays
def prepare_sparse_params(self, param_rowids): '''Prepares the module for processing a data batch by pulling row_sparse parameters from kvstore to all devices based on rowids. Parameters ---------- param_rowids : dict of str to NDArray of list of NDArrays ''' if ...
Saves model parameters to file. Parameters ---------- fname : str Path to output param file. Examples -------- >>> # An example of saving module parameters. >>> mod.save_params('myfile')
def save_params(self, fname): """Saves model parameters to file. Parameters ---------- fname : str Path to output param file. Examples -------- >>> # An example of saving module parameters. >>> mod.save_params('myfile') """ arg_...
Copy data from kvstore to `arg_params` and `aux_params`. Parameters ---------- arg_params : list of NDArray Target parameter arrays. aux_params : list of NDArray Target aux arrays. Notes ----- - This function will inplace update the NDArray...
def get_params_from_kv(self, arg_params, aux_params): """ Copy data from kvstore to `arg_params` and `aux_params`. Parameters ---------- arg_params : list of NDArray Target parameter arrays. aux_params : list of NDArray Target aux arrays. Notes ...
Clips gradient norm. The norm is computed over all gradients together, as if they were concatenated into a single vector. Gradients are modified in-place. The method is first used in `[ICML2013] On the difficulty of training recurrent neural networks` Note that the gradients...
def clip_by_global_norm_per_ctx(self, max_norm=1.0, param_names=None): """Clips gradient norm. The norm is computed over all gradients together, as if they were concatenated into a single vector. Gradients are modified in-place. The method is first used in `[ICML2013] On the ...
Rescale the gradient of provided parameters by a certain scale
def rescale_grad(self, scale=None, param_name=None): """ Rescale the gradient of provided parameters by a certain scale """ if scale is None or param_name is None: return param_idx = self._exec_group.param_names.index(param_name) grad_vals = self._exec_group.grad_arrays[param...
builds factorization machine network with proper formulation: y = w_0 \sum(x_i w_i) + 0.5(\sum\sum<v_i,v_j>x_ix_j - \sum<v_iv_i>x_i^2)
def factorization_machine_model(factor_size, num_features, lr_mult_config, wd_mult_config, init_config): """ builds factorization machine network with proper formulation: y = w_0 \sum(x_i w_i) + 0.5(\sum\sum<v_i,v_j>x_ix_j - \sum<v_iv_i>x_i^2) """ x = mx.symbol.Variable("...
Reshape data into (num_example, batch_size)
def batchify(data, batch_size): """Reshape data into (num_example, batch_size)""" nbatch = data.shape[0] // batch_size data = data[:nbatch * batch_size] data = data.reshape((batch_size, nbatch)).T return data
Tokenizes a text file.
def tokenize(self, path): """Tokenizes a text file.""" assert os.path.exists(path) # Add words to the dictionary with open(path, 'r') as f: tokens = 0 for line in f: words = line.split() + ['<eos>'] tokens += len(words) ...
Build docstring for symbolic functions.
def _build_doc(func_name, desc, arg_names, arg_types, arg_desc, key_var_num_args=None, ret_type=None): """Build docstring for symbolic functions.""" param_str = _build_param_doc(arg_names, arg_types, arg_desc) if key_v...
Get user friendly information of the output shapes.
def get_output_shape(sym, **input_shapes): """Get user friendly information of the output shapes.""" _, s_outputs, _ = sym.infer_shape(**input_shapes) return dict(zip(sym.list_outputs(), s_outputs))
Query CUDA for the number of GPUs present. Raises ------ Will raise an exception on any CUDA error. Returns ------- count : int The number of GPUs.
def num_gpus(): """Query CUDA for the number of GPUs present. Raises ------ Will raise an exception on any CUDA error. Returns ------- count : int The number of GPUs. """ count = ctypes.c_int() check_call(_LIB.MXGetGPUCount(ctypes.byref(count))) return count.value
Query CUDA for the free and total bytes of GPU global memory. Parameters ---------- device_id : int, optional The device id of the GPU device. Raises ------ Will raise an exception on any CUDA error. Returns ------- (free, total) : (int, int) The number of GPUs.
def gpu_memory_info(device_id=0): """Query CUDA for the free and total bytes of GPU global memory. Parameters ---------- device_id : int, optional The device id of the GPU device. Raises ------ Will raise an exception on any CUDA error. Returns ------- (free, total) : ...
Returns the current context. By default, `mx.cpu()` is used for all the computations and it can be overridden by using `with mx.Context(x)` statement where x can be cpu(device_id) or gpu(device_id). Examples ------- >>> mx.current_context() cpu(0) >>> with mx.Context('gpu', 1): # Cont...
def current_context(): """Returns the current context. By default, `mx.cpu()` is used for all the computations and it can be overridden by using `with mx.Context(x)` statement where x can be cpu(device_id) or gpu(device_id). Examples ------- >>> mx.current_context() cpu(0) >>> with...
Populates synsets - a map of index to label for the data items. Populates the data in the dataset, making tuples of (data, label)
def _list_audio_files(self, root, skip_rows=0): """Populates synsets - a map of index to label for the data items. Populates the data in the dataset, making tuples of (data, label) """ self.synsets = [] self.items = [] if not self._train_csv: # The audio files...
Returns a new dataset with the first element of each sample transformed by the transformer function `fn`. This is useful, for example, when you only want to transform data while keeping label as is. lazy=False is passed to transform_first for dataset so that all tramsforms could be perf...
def transform_first(self, fn, lazy=False): """Returns a new dataset with the first element of each sample transformed by the transformer function `fn`. This is useful, for example, when you only want to transform data while keeping label as is. lazy=False is passed to transform_...
Try to configure cython and return cython configuration
def config_cython(): """Try to configure cython and return cython configuration""" if not with_cython: return [] # pylint: disable=unreachable if os.name == 'nt': print("WARNING: Cython is not supported on Windows, will compile without cython module") return [] try: ...
Compose symbol on inputs. This call mutates the current symbol. Parameters ---------- args: provide positional arguments kwargs: provide keyword arguments Returns ------- the resulting symbol
def _compose(self, *args, **kwargs): """Compose symbol on inputs. This call mutates the current symbol. Parameters ---------- args: provide positional arguments kwargs: provide keyword arguments Returns ------- the resul...
Set the attribute of the symbol. Parameters ---------- **kwargs The attributes to set
def _set_attr(self, **kwargs): """Set the attribute of the symbol. Parameters ---------- **kwargs The attributes to set """ keys = c_str_array(kwargs.keys()) vals = c_str_array([str(s) for s in kwargs.values()]) num_args = mx_uint(len(kwargs))...
Configuration factory for various networks Parameters ---------- network : str base network name, such as vgg_reduced, inceptionv3, resnet... data_shape : int input data dimension kwargs : dict extra arguments
def get_config(network, data_shape, **kwargs): """Configuration factory for various networks Parameters ---------- network : str base network name, such as vgg_reduced, inceptionv3, resnet... data_shape : int input data dimension kwargs : dict extra arguments """ ...
Wrapper for get symbol for train Parameters ---------- network : str name for the base network symbol data_shape : int input shape kwargs : dict see symbol_builder.get_symbol_train for more details
def get_symbol_train(network, data_shape, **kwargs): """Wrapper for get symbol for train Parameters ---------- network : str name for the base network symbol data_shape : int input shape kwargs : dict see symbol_builder.get_symbol_train for more details """ if ne...
Set the trainer this parameter is associated with.
def _set_trainer(self, trainer): """ Set the trainer this parameter is associated with. """ # trainer cannot be replaced for sparse params if self._stype != 'default' and self._trainer and trainer and self._trainer is not trainer: raise RuntimeError( "Failed to set th...
Get row_sparse data from row_sparse parameters based on row_id.
def _get_row_sparse(self, arr_list, ctx, row_id): """ Get row_sparse data from row_sparse parameters based on row_id. """ # get row sparse params based on row ids if not isinstance(row_id, ndarray.NDArray): raise TypeError("row_id must have NDArray type, but %s is given"%(type(row_id...
(Re)initializes by loading from data.
def _load_init(self, data, ctx): """(Re)initializes by loading from data.""" if self.shape: for self_dim, data_dim in zip(self.shape, data.shape): assert self_dim in (0, data_dim), \ "Failed loading Parameter '%s' from saved params: " \ ...
Finishes deferred initialization.
def _finish_deferred_init(self): """Finishes deferred initialization.""" if not self._deferred_init: return init, ctx, default_init, data = self._deferred_init self._deferred_init = () assert self.shape is not None and np.prod(self.shape) > 0, \ "Cannot in...
Sets data and grad.
def _init_impl(self, data, ctx_list): """Sets data and grad.""" self._ctx_list = list(ctx_list) self._ctx_map = [[], []] for i, ctx in enumerate(self._ctx_list): dev_list = self._ctx_map[ctx.device_typeid&1] while len(dev_list) <= ctx.device_id: de...
Initialize grad buffers.
def _init_grad(self): """Initialize grad buffers.""" if self.grad_req == 'null': self._grad = None return self._grad = [ndarray.zeros(shape=i.shape, dtype=i.dtype, ctx=i.context, stype=self._grad_stype) for i in self._data] au...
Reduce data from multiple context to cpu.
def _reduce(self): """Reduce data from multiple context to cpu.""" ctx = context.cpu() if self._stype == 'default': block = self.list_data() data = ndarray.add_n(*(w.copyto(ctx) for w in block)) / len(block) else: # fetch all rows for 'row_sparse' para...
Initializes parameter and gradient arrays. Only used for :py:class:`NDArray` API. Parameters ---------- init : Initializer The initializer to use. Overrides :py:meth:`Parameter.init` and default_init. ctx : Context or list of Context, defaults to :py:meth:`context.current_co...
def initialize(self, init=None, ctx=None, default_init=initializer.Uniform(), force_reinit=False): """Initializes parameter and gradient arrays. Only used for :py:class:`NDArray` API. Parameters ---------- init : Initializer The initializer to use. Overrid...
Re-assign Parameter to other contexts. Parameters ---------- ctx : Context or list of Context, default ``context.current_context()``. Assign Parameter to given context. If ctx is a list of Context, a copy will be made for each context.
def reset_ctx(self, ctx): """Re-assign Parameter to other contexts. Parameters ---------- ctx : Context or list of Context, default ``context.current_context()``. Assign Parameter to given context. If ctx is a list of Context, a copy will be made for each context...
Sets this parameter's value on all contexts.
def set_data(self, data): """Sets this parameter's value on all contexts.""" self.shape = data.shape if self._data is None: assert self._deferred_init, \ "Parameter '%s' has not been initialized"%self.name self._deferred_init = self._deferred_init[:3] + (...
Returns a copy of the 'row_sparse' parameter on the same context as row_id's. The copy only retains rows whose ids occur in provided row ids. The parameter must have been initialized on this context before. Parameters ---------- row_id: NDArray Row ids to retain for ...
def row_sparse_data(self, row_id): """Returns a copy of the 'row_sparse' parameter on the same context as row_id's. The copy only retains rows whose ids occur in provided row ids. The parameter must have been initialized on this context before. Parameters ---------- row_...
Returns copies of the 'row_sparse' parameter on all contexts, in the same order as creation. The copy only retains rows whose ids occur in provided row ids. The parameter must have been initialized before. Parameters ---------- row_id: NDArray Row ids to retain for t...
def list_row_sparse_data(self, row_id): """Returns copies of the 'row_sparse' parameter on all contexts, in the same order as creation. The copy only retains rows whose ids occur in provided row ids. The parameter must have been initialized before. Parameters ---------- ...
Returns a copy of this parameter on one context. Must have been initialized on this context before. For sparse parameters, use :py:meth:`Parameter.row_sparse_data` instead. Parameters ---------- ctx : Context Desired context. Returns ------- ...
def data(self, ctx=None): """Returns a copy of this parameter on one context. Must have been initialized on this context before. For sparse parameters, use :py:meth:`Parameter.row_sparse_data` instead. Parameters ---------- ctx : Context Desired context. ...
Returns copies of this parameter on all contexts, in the same order as creation. For sparse parameters, use :py:meth:`Parameter.list_row_sparse_data` instead. Returns ------- list of NDArrays
def list_data(self): """Returns copies of this parameter on all contexts, in the same order as creation. For sparse parameters, use :py:meth:`Parameter.list_row_sparse_data` instead. Returns ------- list of NDArrays """ if self._stype != 'default': ...
Returns a gradient buffer for this parameter on one context. Parameters ---------- ctx : Context Desired context.
def grad(self, ctx=None): """Returns a gradient buffer for this parameter on one context. Parameters ---------- ctx : Context Desired context. """ if self._data is not None and self._grad is None: raise RuntimeError( "Cannot get gr...
Returns gradient buffers on all contexts, in the same order as :py:meth:`values`.
def list_grad(self): """Returns gradient buffers on all contexts, in the same order as :py:meth:`values`.""" if self._data is not None and self._grad is None: raise RuntimeError( "Cannot get gradient array for Parameter '%s' " \ "because grad_req='null...