Add temporal rolling window support - #5393

Open

esnvidia wants to merge 15 commits into

rapidsai:mainfrom

esnvidia:optimization-full-B-C-D-hash

Commits on Jan 13, 2026

Fix null optional dereference in temporal_partition_vertices
esnvidia
committed
ab5104a
Optimization D: Use temporal_sample_edges for inline temporal filtering
esnvidia
committed
cf36570
Optimization B/C: CUDA-level window edge mask primitives
esnvidia
committed
0b2df24
Add B+C+D windowed temporal sampling integration
esnvidia
committed
1fa3806
Add CG-compatible key store and update optimization analysis
esnvidia
committed
f789b09
Add CG-optimized hash table for trillion-edge scale sampling
esnvidia
committed
7e78949
Integrate CG hash table into sampling post-processing
esnvidia
committed
c23641b
Document principled CG size selection from nsys profile analysis
esnvidia
committed
936030b
Fix graph_view.number_of_edges() API for single-GPU graphs
esnvidia
committed
9a9495d
Add pylibcugraph bindings for B+C+D windowed temporal sampling
esnvidia
committed
2a90799
Add flexible timestamp conversion for window parameters
esnvidia
committed
36c89ab
Rename temporal_neighbor_sampling.cpp to .cu for CUDA compilation
esnvidia
committed
4598380
Add comprehensive unit tests for windowed temporal sampling
esnvidia
committed
0a80a55
Windowed temporal sampling: inline window filtering + robustness fixes
esnvidia
committed
f98e6a2
Delete cpp/src/sampling/OPTIMIZATION_PROPOSAL_B_C_D_HASH.md
esnvidia
authored
85fdacc