# Symbol¶

## Get symbol attributes¶

 arguments Get the arguments of symbol children Gets a new grouped symbol whose output contains inputs to output nodes of the original symbol graph.viz Convert symbol to Graphviz or visNetwork visualisation internals Get a symbol that contains all the internals is.mx.symbol Judge if an object is mx.symbol mx.symbol.infer.shape Inference the shape of arguments, outputs, and auxiliary states mx.symbol.shape_array Returns a 1D int64 array containing the shape of data mx.symbol.size_array Returns a 1D int64 array containing the size of data outputs Get the outputs of a symbol

## Create symbols¶

 mx.symbol.Group Create a symbol that groups symbols together mx.symbol.GroupNorm Group normalization mx.symbol.Variable Create a symbolic variable with specified name mx.symbol.one_hot Returns a one-hot array mx.symbol.ones_like Return an array of ones with the same shape and type as the input array mx.symbol.zeros_like Return an array of zeros with the same shape, type and storage type as the input array

## Manipulation of symbols¶

### Conversion¶

 mx.apply Apply symbol to the inputs mx.symbol.cast Casts all elements of the input to a new type mx.symbol.cast_storage Casts tensor storage type to the new type mx.symbol.load Load an mx.symbol object mx.symbol.load.json Load an mx.symbol object from a json string mx.symbol.save Save an mx.symbol object

### Reshaping¶

 mx.symbol.Concat Perform an feature concat on channel dim (dim 1) over all the inputs mx.symbol.Reshape Reshapes the input array mx.symbol.flatten Flattens the input array into a 2-D array by collapsing the higher dimensions mx.symbol.reshape_like Reshape some or all dimensions of lhs to have the same shape as some or all dimensions of rhs mx.symbol.scatter_nd Scatters data into a new tensor according to indices mx.symbol.split Splits an array along a particular axis into multiple sub-arrays mx.symbol.squeeze Remove single-dimensional entries from the shape of an array mx.symbol.stack Join a sequence of arrays along a new axis

### Expanding elements¶

 mx.symbol.GridGenerator Generates 2D sampling grid for bilinear sampling mx.symbol.Pad Pads an input array with a constant or edge values of the array mx.symbol.UpSampling Upsamples the given input data mx.symbol.broadcast_add Returns element-wise sum of the input arrays with broadcasting mx.symbol.broadcast_axes Broadcasts the input array over particular axes mx.symbol.broadcast_axis Broadcasts the input array over particular axes mx.symbol.broadcast_div Returns element-wise division of the input arrays with broadcasting mx.symbol.broadcast_equal Returns the result of element-wise equal to (==) comparison operation with broadcasting mx.symbol.broadcast_greater Returns the result of element-wise greater than (>) comparison operation with broadcasting mx.symbol.broadcast_greater_equal Returns the result of element-wise greater than or equal to (>=) comparison operation with broadcasting mx.symbol.broadcast_hypot Returns the hypotenuse of a right angled triangle, given its “legs” with broadcasting mx.symbol.broadcast_lesser Returns the result of element-wise lesser than (<) comparison operation with broadcasting mx.symbol.broadcast_lesser_equal Returns the result of element-wise lesser than or equal to (<=) comparison operation with broadcasting mx.symbol.broadcast_like Broadcasts lhs to have the same shape as rhs mx.symbol.broadcast_logical_and Returns the result of element-wise logical and with broadcasting mx.symbol.broadcast_logical_or Returns the result of element-wise logical or with broadcasting mx.symbol.broadcast_logical_xor Returns the result of element-wise logical xor with broadcasting mx.symbol.broadcast_maximum Returns element-wise maximum of the input arrays with broadcasting mx.symbol.broadcast_minimum Returns element-wise minimum of the input arrays with broadcasting mx.symbol.broadcast_minus Returns element-wise difference of the input arrays with broadcasting mx.symbol.broadcast_mod Returns element-wise modulo of the input arrays with broadcasting mx.symbol.broadcast_mul Returns element-wise product of the input arrays with broadcasting mx.symbol.broadcast_not_equal Returns the result of element-wise not equal to (!=) comparison operation with broadcasting mx.symbol.broadcast_plus Returns element-wise sum of the input arrays with broadcasting mx.symbol.broadcast_power Returns result of first array elements raised to powers from second array, element-wise with broadcasting mx.symbol.broadcast_sub Returns element-wise difference of the input arrays with broadcasting mx.symbol.broadcast_to Broadcasts the input array to a new shape mx.symbol.expand_dims Inserts a new axis of size 1 into the array shape For example, given x with shape (2,3,4), then expand_dims(x, axis=1) will return a new array with shape (2,1,3,4) mx.symbol.repeat Repeats elements of an array mx.symbol.tile Repeats the whole array multiple times

### Rearranging elements¶

 mx.symbol.SequenceReverse Reverses the elements of each sequence mx.symbol.SwapAxis Interchanges two axes of an array mx.symbol.depth_to_space Rearranges(permutes) data from depth into blocks of spatial data mx.symbol.flip Reverses the order of elements along given axis while preserving array shape mx.symbol.reverse Reverses the order of elements along given axis while preserving array shape mx.symbol.shuffle Randomly shuffle the elements mx.symbol.space_to_depth Rearranges(permutes) blocks of spatial data into depth mx.symbol.swapaxes Interchanges two axes of an array mx.symbol.transpose Permutes the dimensions of an array

### Sorting and searching¶

 mx.symbol.argmax Returns indices of the maximum values along an axis mx.symbol.argmax_channel Returns argmax indices of each channel from the input array mx.symbol.argmin Returns indices of the minimum values along an axis mx.symbol.argsort Returns the indices that would sort an input array along the given axis mx.symbol.sort Returns a sorted copy of an input array along the given axis mx.symbol.topk Returns the indices of the top k elements in an input array along the given axis (by default) mx.symbol.where Return the elements, either from x or y, depending on the condition

### Indexing¶

 mx.symbol.Crop mx.symbol.SequenceLast Takes the last element of a sequence mx.symbol.SequenceMask Sets all elements outside the sequence to a constant value mx.symbol.SliceChannel Splits an array along a particular axis into multiple sub-arrays mx.symbol.batch_take Takes elements from a data batch mx.symbol.choose_element_0index Picks elements from an input array according to the input indices along the given axis mx.symbol.diag Extracts a diagonal or constructs a diagonal array mx.symbol.fill_element_0index Fill one element of each line(row for python, column for R/Julia) in lhs according to index indicated by rhs and values indicated by mhs mx.symbol.gather_nd Gather elements or slices from data and store to a tensor whose shape is defined by indices mx.symbol.pick Picks elements from an input array according to the input indices along the given axis mx.symbol.ravel_multi_index Converts a batch of index arrays into an array of flat indices mx.symbol.slice Slices a region of the array mx.symbol.slice_axis Slices along a given axis mx.symbol.slice_like Slices a region of the array like the shape of another array mx.symbol.take Takes elements from an input array along the given axis mx.symbol.unravel_index Converts an array of flat indices into a batch of index arrays

## Mathematical operations on symbols¶

### Arithmetic¶

 mx.symbol.ElementWiseSum Adds all input arguments element-wise mx.symbol.abs Returns element-wise absolute value of the input mx.symbol.add_n Adds all input arguments element-wise mx.symbol.all_finite Check if all the float numbers in the array are finite (used for AMP) mx.symbol.cbrt Returns element-wise cube-root value of the input mx.symbol.cumsum Return the cumulative sum of the elements along a given axis mx.symbol.elemwise_add Adds arguments element-wise mx.symbol.elemwise_div Divides arguments element-wise mx.symbol.elemwise_mul Multiplies arguments element-wise mx.symbol.elemwise_sub Subtracts arguments element-wise mx.symbol.erf Returns element-wise gauss error function of the input mx.symbol.erfinv Returns element-wise inverse gauss error function of the input mx.symbol.exp Returns element-wise exponential value of the input mx.symbol.expm1 Returns exp(x) - 1 computed element-wise on the input mx.symbol.gamma Returns the gamma function (extension of the factorial function to the reals), computed element-wise on the input array mx.symbol.gammaln Returns element-wise log of the absolute value of the gamma function of the input mx.symbol.identity Returns a copy of the input mx.symbol.log Returns element-wise Natural logarithmic value of the input mx.symbol.log10 Returns element-wise Base-10 logarithmic value of the input mx.symbol.log1p Returns element-wise log(1 + x) value of the input mx.symbol.log2 Returns element-wise Base-2 logarithmic value of the input mx.symbol.logical_not Returns the result of logical NOT (!) function mx.symbol.moments Calculate the mean and variance of data mx.symbol.multi_all_finite Check if all the float numbers in all the arrays are finite (used for AMP) mx.symbol.multi_lars Compute the LARS coefficients of multiple weights and grads from their sums of square” mx.symbol.multi_sum_sq Compute the sums of squares of multiple arrays mx.symbol.negative Numerical negative of the argument, element-wise mx.symbol.rcbrt Returns element-wise inverse cube-root value of the input mx.symbol.reciprocal Returns the reciprocal of the argument, element-wise mx.symbol.reset_arrays Set to zero multiple arrays mx.symbol.rsqrt Returns element-wise inverse square-root value of the input mx.symbol.sign Returns element-wise sign of the input mx.symbol.sqrt Returns element-wise square-root value of the input mx.symbol.square Returns element-wise squared value of the input

### Reduce¶

 mx.symbol.max Computes the max of array elements over given axes mx.symbol.max_axis Computes the max of array elements over given axes mx.symbol.mean Computes the mean of array elements over given axes mx.symbol.nanprod Computes the product of array elements over given axes treating Not a Numbers (NaN) as one mx.symbol.nansum Computes the sum of array elements over given axes treating Not a Numbers (NaN) as zero mx.symbol.prod Computes the product of array elements over given axes mx.symbol.sum Computes the sum of array elements over given axes mx.symbol.sum_axis Computes the sum of array elements over given axes

### Round¶

 mx.symbol.ceil Returns element-wise ceiling of the input mx.symbol.clip Clips (limits) the values in an array mx.symbol.fix Returns element-wise rounded value to the nearest integer towards zero of the input mx.symbol.floor Returns element-wise floor of the input mx.symbol.rint Returns element-wise rounded value to the nearest integer of the input mx.symbol.round Returns element-wise rounded value to the nearest integer of the input mx.symbol.trunc Return the element-wise truncated value of the input

### Linear algebra¶

 mx.symbol.L2Normalization Normalize the input array using the L2 norm mx.symbol.batch_dot Batchwise dot product mx.symbol.dot Dot product of two arrays mx.symbol.khatri_rao Computes the Khatri-Rao product of the input matrices mx.symbol.linalg_det Compute the determinant of a matrix mx.symbol.linalg_extractdiag Extracts the diagonal entries of a square matrix mx.symbol.linalg_extracttrian Extracts a triangular sub-matrix from a square matrix mx.symbol.linalg_gelqf LQ factorization for general matrix mx.symbol.linalg_gemm Performs general matrix multiplication and accumulation mx.symbol.linalg_gemm2 Performs general matrix multiplication mx.symbol.linalg_inverse Compute the inverse of a matrix mx.symbol.linalg_makediag Constructs a square matrix with the input as diagonal mx.symbol.linalg_maketrian Constructs a square matrix with the input representing a specific triangular sub-matrix mx.symbol.linalg_potrf Performs Cholesky factorization of a symmetric positive-definite matrix mx.symbol.linalg_potri Performs matrix inversion from a Cholesky factorization mx.symbol.linalg_slogdet Compute the sign and log of the determinant of a matrix mx.symbol.linalg_sumlogdiag Computes the sum of the logarithms of the diagonal elements of a square matrix mx.symbol.linalg_syrk Multiplication of matrix with its transpose mx.symbol.linalg_trmm Performs multiplication with a lower triangular matrix mx.symbol.linalg_trsm Solves matrix equation involving a lower triangular matrix mx.symbol.norm Computes the norm on an NDArray mx.symbol.smooth_l1 Calculate Smooth L1 Loss(lhs, scalar) by summing

### Trigonometric functions¶

 mx.symbol.arccos Returns element-wise inverse cosine of the input array mx.symbol.arcsin Returns element-wise inverse sine of the input array mx.symbol.arctan Returns element-wise inverse tangent of the input array mx.symbol.cos Computes the element-wise cosine of the input array mx.symbol.degrees Converts each element of the input array from radians to degrees mx.symbol.radians Converts each element of the input array from degrees to radians mx.symbol.sin Computes the element-wise sine of the input array mx.symbol.tan Computes the element-wise tangent of the input array

### Hyperbolic functions¶

 mx.symbol.arccosh Returns the element-wise inverse hyperbolic cosine of the input array, computed element-wise mx.symbol.arcsinh Returns the element-wise inverse hyperbolic sine of the input array, computed element-wise mx.symbol.arctanh Returns the element-wise inverse hyperbolic tangent of the input array, computed element-wise mx.symbol.cosh Returns the hyperbolic cosine of the input array, computed element-wise mx.symbol.sinh Returns the hyperbolic sine of the input array, computed element-wise mx.symbol.tanh Returns the hyperbolic tangent of the input array, computed element-wise

## Neural network symbol operations¶

 mx.symbol.Activation Applies an activation function element-wise to the input mx.symbol.BatchNorm Batch normalization mx.symbol.BatchNorm_v1 Batch normalization mx.symbol.BlockGrad Stops gradient computation mx.symbol.CTCLoss Connectionist Temporal Classification Loss mx.symbol.Convolution Compute N-D convolution on (N+2)-D input mx.symbol.Convolution_v1 This operator is DEPRECATED mx.symbol.Correlation Applies correlation to inputs mx.symbol.Custom Apply a custom operator implemented in a frontend language (like Python) mx.symbol.Deconvolution Computes 1D or 2D transposed convolution (aka fractionally strided convolution) of the input tensor mx.symbol.Dropout Applies dropout operation to input array mx.symbol.Embedding Maps integer indices to vector representations (embeddings) mx.symbol.FullyConnected Applies a linear transformation: $$Y = XW^T + b$$ mx.symbol.IdentityAttachKLSparseReg Apply a sparse regularization to the output a sigmoid activation function mx.symbol.InstanceNorm Applies instance normalization to the n-dimensional input array mx.symbol.LRN Applies local response normalization to the input mx.symbol.LayerNorm Layer normalization mx.symbol.LeakyReLU Applies Leaky rectified linear unit activation element-wise to the input mx.symbol.LinearRegressionOutput Computes and optimizes for squared loss during backward propagation mx.symbol.LogisticRegressionOutput Applies a logistic function to the input mx.symbol.MAERegressionOutput Computes mean absolute error of the input mx.symbol.MakeLoss Make your own loss function in network construction mx.symbol.Pooling Performs pooling on the input mx.symbol.Pooling_v1 This operator is DEPRECATED mx.symbol.RNN Applies recurrent layers to input data mx.symbol.ROIPooling Performs region of interest(ROI) pooling on the input array mx.symbol.SVMOutput Computes support vector machine based transformation of the input mx.symbol.SoftmaxActivation Applies softmax activation to input mx.symbol.SoftmaxOutput Computes the gradient of cross entropy loss with respect to softmax output mx.symbol.SpatialTransformer Applies a spatial transformer to input feature map mx.symbol.ctc_loss Connectionist Temporal Classification Loss mx.symbol.hard_sigmoid Computes hard sigmoid of x element-wise mx.symbol.log_softmax Computes the log softmax of the input mx.symbol.make_loss Make your own loss function in network construction mx.symbol.relu Computes rectified linear activation mx.symbol.sigmoid Computes sigmoid of x element-wise mx.symbol.softmax Applies the softmax function mx.symbol.softmax_cross_entropy Calculate cross entropy of softmax output and one-hot label mx.symbol.softmin Applies the softmin function mx.symbol.softsign Computes softsign of x element-wise mx.symbol.stop_gradient Stops gradient computation