Numpy Interfaces in MindSpore

Overview

MindSpore Numpy package contains a set of Numpy-like interfaces, which allows developers to build models on MindSpore with similar syntax of Numpy.

Operator Functions

Mindspore Numpy operators can be classified into four functional modules: array generation, array operation, logic operation and math operation. For details about the supported operators on the Ascend AI processors, GPU, and CPU, see Numpy Interface List.

Array Generations

Array generation operators are used to generate tensors.

Here is an example to generate an array:

import mindspore.numpy as np
import mindspore.ops as ops
input_x = np.array([1, 2, 3], np.float32)
print("input_x =", input_x)
print("type of input_x =", ops.typeof(input_x))

The output is:

input_x = [1. 2. 3.]
type of input_x = Tensor[Float32]

Here we have more examples:

Generate a tensor filled with the same element

np.full can be used to generate a tensor with user-specified values:

import mindspore.numpy as np
input_x = np.full((2, 3), 6, np.float32)
print(input_x)

The output is:

[[6. 6. 6.]
 [6. 6. 6.]]

Here is another example to generate an array with the specified shape and filled with the value of 1:

import mindspore.numpy as np
input_x = np.ones((2, 3), np.float32)
print(input_x)

The output is:

[[1. 1. 1.]
 [1. 1. 1.]]

Generate tensors in a specified range

Generate an arithmetic array within the specified range：

import mindspore.numpy as np
input_x = np.arange(0, 5, 1)
print(input_x)

The output is:

[0 1 2 3 4]

Generate tensors with specific requirement

Generate a matrix where the lower elements are 1 and the upper elements are 0 on the given diagonal:

import mindspore.numpy as np
input_x = np.tri(3, 3, 1)
print(input_x)

The output is:

[[1. 1. 0.]
 [1. 1. 1.]
 [1. 1. 1.]]

Another example, generate a 2-D matrix with a diagonal of 1 and other elements of 0:

import mindspore.numpy as np
input_x = np.eye(2, 2)
print(input_x)

The output is:

[[1. 0.]
 [0. 1.]]

Array Operations

Array operations focus on tensor manipulation.

Manipulate the shape of the tensor

For example, transpose a matrix:

import mindspore.numpy as np
input_x = np.arange(10).reshape(5, 2)
output = np.transpose(input_x)
print(output)

The output is:

[[0 2 4 6 8]
 [1 3 5 7 9]]

Another example, swap two axes:

import mindspore.numpy as np
input_x = np.ones((1, 2, 3))
output = np.swapaxes(input_x, 0, 1)
print(output.shape)

The output is:

(2, 1, 3)

Tensor splitting

Divide the input tensor into multiple tensors equally, for example:

import mindspore.numpy as np
input_x = np.arange(9)
output = np.split(input_x, 3)
print(output)

The output is:

(Tensor(shape=[3], dtype=Int32, value= [0, 1, 2]),
 Tensor(shape=[3], dtype=Int32, value= [3, 4, 5]),
 Tensor(shape=[3], dtype=Int32, value= [6, 7, 8]))

Tensor combination

Concatenate the two tensors according to the specified axis, for example:

import mindspore.numpy as np
input_x = np.arange(0, 5)
input_y = np.arange(10, 15)
output = np.concatenate((input_x, input_y), axis=0)
print(output)

The output is:

[ 0  1  2  3  4 10 11 12 13 14]

Logic Operations

Logic operations define computations related with boolean types. Examples of equal and less operations are as follows:

import mindspore.numpy as np
input_x = np.arange(0, 5)
input_y = np.arange(0, 10, 2)
output = np.equal(input_x, input_y)
print("output of equal:", output)
output = np.less(input_x, input_y)
print("output of less:", output)

The output is:

output of equal: [ True False False False False]
output of less: [False  True  True  True  True]

Math Operations

Math operations include basic and advanced math operations on tensors, and they have full support on Numpy broadcasting rules. Here are some examples:

Sum two tensors

The following code implements the operation of adding two tensors of input_x and input_y:

import mindspore.numpy as np
input_x = np.full((3, 2), [1, 2])
input_y = np.full((3, 2), [3, 4])
output = np.add(input_x, input_y)
print(output)

The output is:

[[4 6]
 [4 6]
 [4 6]]

Matrics multiplication

The following code implements the operation of multiplying two matrices input_x and input_y:

import mindspore.numpy as np
input_x = np.arange(2*3).reshape(2, 3).astype('float32')
input_y = np.arange(3*4).reshape(3, 4).astype('float32')
output = np.matmul(input_x, input_y)
print(output)

The output is:

[[20. 23. 26. 29.]
 [56. 68. 80. 92.]]

Take the average along a given axis

The following code implements the operation of averaging all the elements of input_x:

import mindspore.numpy as np
input_x = np.arange(6).astype('float32')
output = np.mean(input_x)
print(output)

The output is:

2.5

Exponential arithmetic

The following code implements the operation of the natural constant e to the power of input_x:

import mindspore.numpy as np
input_x = np.arange(5).astype('float32')
output = np.exp(input_x)
print(output)

The output is:

[ 1.         2.718282   7.3890557 20.085537  54.598145 ]

Interact With MindSpore Functions

Since mindspore.numpy directly wraps MindSpore tensors and operators, it has all the advantages and properties of MindSpore. In this section, we will briefly introduce how to employ MindSpore execution management and automatic differentiation in mindspore.numpy coding scenarios. These include:

ms_function: for running codes in static graph mode for better efficiency.
GradOperation: for automatic gradient computation.
mindspore.context: for mindspore.numpy execution management.
mindspore.nn.Cell: for using mindspore.numpy interfaces in MindSpore Deep Learning Models.

Use ms_function to run code in static graph mode

Let’s first see an example consisted of matrix multiplication and bias add, which is a typical process in Neural Networks:

import mindspore.numpy as np

x = np.arange(8).reshape(2, 4).astype('float32')
w1 = np.ones((4, 8))
b1 = np.zeros((8,))
w2 = np.ones((8, 16))
b2 = np.zeros((16,))
w3 = np.ones((16, 4))
b3 = np.zeros((4,))

def forward(x, w1, b1, w2, b2, w3, b3):
    x = np.dot(x, w1) + b1
    x = np.dot(x, w2) + b2
    x = np.dot(x, w3) + b3
    return x

print(forward(x, w1, b1, w2, b2, w3, b3))

The output is:

[[ 768.  768.  768.  768.]
 [2816. 2816. 2816. 2816.]]

In this function, MindSpore dispatches each computing kernel to device separately. However, with the help of ms_function, we can compile all operations into a single static computing graph.

from mindspore import ms_function

forward_compiled = ms_function(forward)

Currently, static graph cannot run in command line mode and not all python types can be passed into functions decorated with ms_function. For details about the static graph syntax support, see Syntax Support. For details about how to use ms_function, see API: ms_function.

Use GradOperation to compute deratives

GradOperation can be used to take deratives from normal functions and functions decorated with ms_function. Take the previous example:

from mindspore import ops

grad_all = ops.composite.GradOperation(get_all=True)
grad_all(forward)(x, w1, b1, w2, b2, w3, b3)

To take the gradient of ms_function compiled functions, first we need to set the execution mode to static graph mode.

from mindspore import ops, ms_function, context

context.set_context(mode=context.GRAPH_MODE)

grad_all = ops.composite.GradOperation(get_all=True)
grad_all(ms_function(forward))(x, w1, b1, w2, b2, w3, b3)

For more details, see API: GradOperation.

Use mindspore.context to control execution mode

Most functions in mindspore.numpy can run in Graph Mode and PyNative Mode, and can run on CPU，GPU and Ascend. Like MindSpore, users can manage the execution mode using mindspore.context：

import mindspore.numpy as np
from mindspore import context

# Execution in static graph mode
context.set_context(mode=context.GRAPH_MODE)

# Execution in dynamic graph mode
context.set_context(mode=context.PYNATIVE_MODE)

# Execution on CPU backend
context.set_context(device_target="CPU")

# Execution on GPU backend
context.set_context(device_target="GPU")

# Execution on Ascend backend
context.set_context(device_target="Ascend")
...

For more details, see API: mindspore.context.

Use mindspore.numpy in MindSpore Deep Learning Models

mindspore.numpy interfaces can be used inside nn.cell blocks as well. For example, the above code can be modified to:

import mindspore.numpy as np
from mindspore import context
from mindspore.nn import Cell

context.set_context(mode=context.GRAPH_MODE)

x = np.arange(8).reshape(2, 4).astype('float32')
w1 = np.ones((4, 8))
b1 = np.zeros((8,))
w2 = np.ones((8, 16))
b2 = np.zeros((16,))
w3 = np.ones((16, 4))
b3 = np.zeros((4,))

class NeuralNetwork(Cell):
    def __init__(self):
        super(NeuralNetwork, self).__init__()
    def construct(self, x, w1, b1, w2, b2, w3, b3):
        x = np.dot(x, w1) + b1
        x = np.dot(x, w2) + b2
        x = np.dot(x, w3) + b3
        return x

net = NeuralNetwork()

print(net(x, w1, b1, w2, b2, w3, b3))

The output is:

[[ 768.  768.  768.  768.]
 [2816. 2816. 2816. 2816.]]

For more details on building Neural Network with MindSpore, see MindSpore Training Guide.