PometryDocs
GitHub

Performance Optimization

Tune Raphtory for your workload

Optimize graph building, algorithms, and queries.

Graph Building Optimization

Batch Loading

Pre-sort Data

Algorithm Optimization

Use Sampling for Large Graphs

Reduce Iterations

Temporal Windows

Memory Optimization

Clear Unused Results

Stream Processing

Parallelization

Set thread count:

Query Optimization

Cache Algorithm Results

Profiling

Find bottlenecks:

Performance Checklist

  • Use load_edges_from_pandas() not row-by-row
  • Sort data by time before ingestion
  • Use sampling for large-graph algorithms
  • Reduce iterations where acceptable
  • Analyze temporal windows, not full history
  • Cache algorithm results
  • Set RAYON_NUM_THREADS appropriately
  • Profile to find actual bottlenecks

See Also