TopoBench

TopoBench copied to clipboard

adding funding

1

Adding funding sources to the readme file, please add yours if any.

mhajij

vapavlo

vapavlo

[1: topobenchmarkx/io/load/utils.py](https://github.com/pyt-team/TopoBenchmarkX/blob/main/topobenchmarkx/io/load/utils.py#L18) [2: topobenchmarkx/io/load/utils.py](https://github.com/pyt-team/TopoBenchmarkX/blob/main/topobenchmarkx/io/load/utils.py#L92) Temporally added "PERF203" to Ruff ignores

vapavlo

Some TopoModelX models do not have normalization of outputs, which results in an inability to stack multiple layers. For example, now we use external implementation of cwn, can be found...

levtelyatnikov

fix formatting error in reproducing scripts

mathildepapillon

Currently the data lifting process produces both up and down laplacian neighborhoods, only up adjacency (not down) neighborhood. For instance, the neighborhoods for a complex are: - up_laplacian_0 - up_laplacian_1...

mathildepapillon

Facilitate the implementation of models that need custom losses --right now the loss is mainly defined in dataset config files.

gbg141

Running Zinc needs adding the command `dataset.transforms.one_hot_node_degree_features.degrees_fields=x` to properly generate its features. This can be automated within `transforms/dataset_defaults/ZINC.yaml`.

gbg141

Added batching in transductive setting

2

Hello everyone, I have added the possibility of batching the data in the transductive setting. When working with large graphs, selecting a subset of the graph while keeping the model's...

Coerulatus