Paddle list index/slice alternative method

[DrRyanHuang]

需求描述 Feature Description

Hi, thx for good jobs of Paddlepaddle!

I need an alternative method of paddle list slice/index. Could you please give me some advices?

numpy implementation:

import numpy as np
t_np = np.random.random([32, 7, 2, 40, 100])
t_np[1, :, :, [1, 2,-1], [1, 3,-1]].shape

(3, 7, 2)

torch implementation：

t_np = np.random.random([32, 7, 2, 40, 100])
t_tc = torch.Tensor(t_np)
t_tc[1, :, :, [1, 2,-1], [1, 3,-1]].shape

[7, 2, 3])

However

import paddle

t_pd = paddle.to_tensor(t_np)
t_pd[1, :, :, [1, 2,-1], [1, 3,-1]].shape

it raise:

(InvalidArgument) When index contains a list, its length is excepted to 1, but received 5

is there any simple alternative method for above implementation in paddle?

替代实现 Alternatives

No response

[paddle-bot]

您好，我们已经收到了您的问题，会安排技术人员尽快解答您的问题，请耐心等待。请您再次检查是否提供了清晰的问题描述、复现代码、环境&版本、报错信息等。同时，您也可以通过查看官网API文档、常见问题、历史Issue、AI社区来寻求解答。祝您生活愉快～

Hi! We've received your issue and please be patient to get responded. We will arrange technicians to answer your questions as soon as possible. Please make sure that you have posted enough message to demo your request. You may also check out the API，FAQ，Github Issue and AI community to get the answer.Have a nice day!

[huangjun12]

import numpy as np
import paddle

t_np = np.random.random([32, 7, 2, 40, 100])
t_pd = paddle.to_tensor(t_np)

a = t_pd[1, :,:,:,:]

tmp_list = [a[:,:, 1, 1].unsqueeze(0), a[:,:, 2, 3].unsqueeze(0), a[:,:, -1, -1].unsqueeze(0)]

res = paddle.concat(tmp_list, axis=0)
print(res.shape)

with paddle develop, this code works well

[DrRyanHuang]

@huangjun12 Thx!