Add DGraphFin dynamic graph dataset

[EdisonLeeeee]

This PR is a re-opened one of #5466. The major difference it that this PR enforces manual download and just provides a process script for the dataset.

• Home page:https://dgraph.xinye.com/introduction
• Paper: https://openreview.net/forum?id=2rQPxsmjKF&referrer=%5Bthe%20
• Baselines: https://github.com/DGraphXinye/DGraphFin_baseline

Usage

from torch_geometric.datasets import DGraphFin
dataset = DGraphFin("data/DGraphFin")
data = dataset[0]
>>> data
Data(x=[3700550, 17], edge_index=[2, 4300999], edge_attr=[4300999], y=[3700550], edge_time=[4300999], train_mask=[3700550], val_mask=[3700550], test_mask=[3700550])

For now, I used Data instead of TemporalData for storing this dataset.

cc: @Padarn @rusty1s

[codecov]

Codecov Report

Merging #5504 (68ed027) into master (d8d06e1) will decrease coverage by 1.98%.
The diff coverage is n/a.

❗ Current head 68ed027 differs from pull request most recent head 4033475. Consider uploading reports for the commit 4033475 to get more accurate results

@@            Coverage Diff             @@
##           master    #5504      +/-   ##
==========================================
- Coverage   85.76%   83.78%   -1.99%     
==========================================
  Files         346      346              
  Lines       19002    18998       -4     
==========================================
- Hits        16298    15917     -381     
- Misses       2704     3081     +377     

Impacted Files	Coverage Δ
torch_geometric/nn/models/dimenet_utils.py	0.00% <0.00%> (-75.52%)	⬇️
torch_geometric/nn/models/dimenet.py	14.51% <0.00%> (-53.00%)	⬇️
torch_geometric/utils/scatter.py	66.66% <0.00%> (-33.34%)	⬇️
torch_geometric/profile/profile.py	37.89% <0.00%> (-26.32%)	⬇️
torch_geometric/nn/conv/utils/typing.py	81.25% <0.00%> (-17.50%)	⬇️
torch_geometric/nn/pool/asap.py	92.10% <0.00%> (-7.90%)	⬇️
torch_geometric/nn/inits.py	67.85% <0.00%> (-7.15%)	⬇️
torch_geometric/transforms/add_self_loops.py	94.44% <0.00%> (-5.56%)	⬇️
torch_geometric/nn/resolver.py	88.88% <0.00%> (-5.56%)	⬇️
torch_geometric/sampler/neighbor_sampler.py	92.90% <0.00%> (-5.17%)	⬇️
... and 11 more

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[wsad1]

Mostly looks good. Left some initial comments.

[wsad1]

Thanks for adding this.

[EdisonLeeeee]

Shall I add an example for the use of this dataset?

[EdisonLeeeee]

Let's leave the example as the follow-up PR if it is necessary. I think this PR is ready to merge.

[rusty1s]

Thank you!