問題文
A team needs to run five different algorithms over the same set of nodes and relationships. What is the efficient approach?
選択肢
- Project the graph before each call so that every algorithm reads the freshest data that is available in the database at that point.
- Project five graphs under different names, one per algorithm, so that the results written by each of them do not overwrite each other in the projection.
- Run all five in stats mode first and then decide from the summaries which of them to project a graph for later.
- Project the graph once and refer to it by name in each of the five calls that come after it, since projecting the same data again for each call would read the database five times over.