dmlc/dgl
Ver no GitHub[Feature Request] Add missing functions in numpy backend.
Open
#664 aberto em 17 de jun. de 2019
help wanted
Métricas do repositório
- Stars
- (12.665 stars)
- Métricas de merge de PR
- (Nenhuma PRs mesclada em 30d)
Description
🚀 Feature
There are some missing or incompatible functions compared with the interface in numpy backend. The list is following.
Missing
- device_type
- mean
- unsorted_1d_segment_sum
- unsorted_1d_segment_mean
- zerocopy_to_dgl_ndarray
- logical_not
- zerocopy_to_dlpack
- equal
- ndim
- sync
- zerocopy_from_dlpack
- copy_reduce
- zerocopy_from_dgl_ndarray
- boolean_mask
- narrow_row
- stack
- binary_reduce
- device_id
- zeros_like
Not compatible
- zeros
(shape, dtype)v.s.(shape, dtype, ctx) - ones
(shape, dtype)v.s.(shape, dtype, ctx) - full_1d
(length, fill_value)v.s.(length, fill_value, dtype, ctx)
Motivation
In some scenarios, we just want to use message propogation in dgl (e.g. LPA, PageRank) and thus choose numpy as backend.
However due to the missing or incompatible functions in numpy backend, even the simplest example send(message_func=dgl.function.copy_src("f", "m")) will raise an exception.
Do you think it's necessary to completing the missing functions? And if so I can help to add some within my ability.