mindspore.ops.linspace

mindspore.ops.linspace(start, stop, num)

Returns a Tensor whose value is num evenly spaced in the interval start and stop (including start and stop), and the length of the output Tensor is num.

$$\begin{array}{r}\begin{array}{rl}& step=(stop-start)/(num-1)\\ & output=[start,start+step,start+2\ast step,...,stop]\end{array}\end{array}$$

Parameters

• start (Tensor) – Start value of interval. The tensor data type must be float32 or float64 and with shape of 0-D.

• stop (Tensor) – Last value of interval. The tensor data type must be float32 or float64 and with shape of 0-D.

• num (Union[Tensor, int]) – Number of ticks in the interval, inclusive of start and stop. Must be positive int number or 0D int32/int64 Tensor.

Returns

Tensor, has the same dtype as start, and the shape of $(num)$.

Raises

• TypeError – If start or stop is not a Tensor.

• TypeError – If dtype of start or dtype of stop is not float32 or float64.

• ValueError – If shape of start or shape of stop is not 0-D.

• TypeError – If num is not int or 0D int32/int64 Tensor.

• ValueError – If num is not positive int number.

Supported Platforms:

Ascend GPU CPU

Examples

>>> start = Tensor(1, mindspore.float32)
>>> stop = Tensor(10, mindspore.float32)
>>> num = 5
>>> output = ops.linspace(start, stop, num)
>>> print(output)
[ 1.    3.25  5.5   7.75 10.  ]