Hi! Macarse wrote: > 1) > Emulating a count(*). > I did this: http://pastebin.com/f276fd2c7 but it takes some time and > it doesn't look like a good solution. > Which is the best way? >
The code uses traversals to infinite depth, which of course comes at a cost even when using Neo4j (and I think this has very little in common with count(*) which only counts to depth one!). How many nodes are typically traversed by your code? How much time is spent per node in the traversal? > Does the following make any sense? > Robert connected to the place and therefore to people living there. > (Robert <--- friend with people living in ---> Argentina; Charly <--- > lives in ---> Argentina) > Makes sense to me. One detail: make sure to use directed edges where it fits your domain when presenting this stuff! :-) > 3) > Is there a way to create something similar to an schema to generate the db? > Robert has almost 100k friends :( > Then Robert has to be a very nice guy ;-) If you have the data in a RDBMS you should take a look at the SQL importer being developed right now: http://wiki.neo4j.org/content/Neo_-_SQL_importer /anders _______________________________________________ Neo mailing list [email protected] https://lists.neo4j.org/mailman/listinfo/user
