問題文
A team wants to detect communities only inside each connected piece of a large graph. They plan to run a components algorithm first and then a community detection algorithm. Which mode should the first algorithm use so the second can build on its output?
選択肢
- The stats mode, which is the basis of the other modes and therefore carries the result forward to the call that follows.
- The write mode, because only values in the database can be read by another algorithm afterwards.
- The mutate mode, so the component identifier is available on the projected graph for the next call to read.
- The stream mode, because the returned rows are passed straight into the second algorithm as its input by the query that calls them in sequence.