Skip to content

Conversation

@DesmonDay
Copy link
Contributor

@DesmonDay DesmonDay commented Sep 20, 2022

Add docs for paddle.geometric APIs.

@paddle-bot
Copy link

paddle-bot bot commented Sep 20, 2022

感谢你贡献飞桨文档,文档预览构建中,Docs-New 跑完后即可预览,预览链接:http://preview-pr-5292.paddle-docs-preview.paddlepaddle.org.cn/documentation/docs/zh/api/index_cn.html
预览工具的更多说明,请参考:[Beta]飞桨文档预览工具


.. math::

out_i = \max_{j \in \{segment\_ids_j == i \} } data_{j}
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

公式预览是这样,是否有问题?
image

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

这个没有问题,确实是这样

参数
:::::::::
- **data** (Tensor) - 张量,数据类型为 float32、float64。
- **segment_ids** (Tensor) - 一维张量,与输入数据`data`的第一维大小相同,表示`data`分段位置,单调非减。合法的数据类型为 int32、int64。
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

data 这里,或许是 data

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

``双引号


.. math::

out_i = \mathop{mean}_{j \in \{segment\_ids_j == i \} } data_{j}
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

同,公式有问题不?
image

参数
:::::::::
- **data** (Tensor) - 张量,数据类型为 float32、float64。
- **segment_ids** (Tensor) - 一维张量,与输入数据`data`的第一维大小相同,表示`data`分段位置,单调非减。合法的数据类型为 int32、int64。
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

``


.. math::

out_i = \min_{j \in \{segment\_ids_j == i \} } data_{j}
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

同公式,下标太长,有点


参数
:::::::::
- x (Tensor) - 输入的节点特征Tensor,数据类型为:float32、float64、int32、int64。另外,我们在 GPU 计算中支持 float16。
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

** 加粗,前4个参数


参数
:::::::::
- x (Tensor) - 输入的节点特征Tensor,数据类型为:float32、float64、int32、int64。另外,我们在 GPU 计算中支持 float16。
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

加粗**

:::::::::
- x (Tensor) - 输入的节点特征Tensor,数据类型为:float32、float64、int32、int64。另外,我们在 GPU 计算中支持 float16。
- src_index (Tensor) - 1-D Tensor,数据类型为:int32、int64。
- dst_index (Tensor) - 1-D Tensor,数据类型为:int32、int64。注意:``dst_index`` 的形状应当与 ``src_index`` 一致。
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

加粗


参数
:::::::::
- x (Tensor) - 输入的中心节点原始编号,数据类型为:int32、int64。
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

参数这里,都加粗一下~

Copy link
Collaborator

@momozi1996 momozi1996 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

一点点小问题

@DesmonDay
Copy link
Contributor Author

除公式问题不做修改外,其余均已修改。


" :ref:`paddle.geometric.send_u_recv <cn_api_paddle_geometric_send_u_recv>` ", "源节点特征消息传递"
" :ref:`paddle.geometric.send_ue_recv <cn_api_paddle_geometric_send_ue_recv>` ", "节点融合边特征消息传递"
" :ref:`paddle.geometric.send_uv <cn_api_paddle_geometric_send_uv>` ", "节点特征消息发送"
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

这里和上面源节点特征消息传递的区别没有特别说清楚

Copy link
Collaborator

@momozi1996 momozi1996 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM


主要应用于图学习领域,需要与图采样相关的 API 配合使用,主要处理同构图场景。其主要目的是对输入的中心节点信息和邻居信息进行从 0 开始的重新编号,以方便后续的图模型子图训练。

**注意**:输入 ``x`` 中的元素需保证是独有的,否则可能会带来一些潜在的错误。输入的节点将会和邻居节点一同从0进行编号。
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

done

@DesmonDay DesmonDay merged commit fab7bcb into PaddlePaddle:develop Oct 11, 2022
DesmonDay added a commit to DesmonDay/docs that referenced this pull request Oct 11, 2022
jzhang533 pushed a commit that referenced this pull request Oct 25, 2022
jzhang533 pushed a commit that referenced this pull request Nov 21, 2022
* Add paddle.geometric docs (#5292)

* add paddle.geometric

* [geometric] delete sampling and message_passing directory (#5427)
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants