I've done some reading about SQLite. Can someone confirm if it is suitable for personal database which require a lot of calculation, analsis and graph? Currently I am working on Excel on 63 files sizing of 1.34GB. I will be creating a million row and maybe 45+ column depend on which table.
A:
I've been working with SQLite extensively during the past six months, and I can tell you that [pshhhsttt message garbled. Entropy critically low due to poor accept rate.]
Calvin1602
2010-08-24 08:21:45
Much as I agree with the sentiment, let's reserve answers for answers.
David Dorward
2010-08-24 08:45:49
Argh. I would have edited the answer if OP had done something... The goal was to prevent other people from answering before this (but now he has his answer, so...). Anyway, no hard feelings :)
Calvin1602
2010-08-24 09:28:44
Please, let's keep the meta-discussion to http://meta.stackoverflow.com/ , or in unavoidable cases to a comment. (yes, the OP should do something about hir accept rate, but the answer should be relevant to question, not the questioner)
Piskvor
2010-08-24 15:02:45
+1
A:
Many people report that SQLite works well for databases sized up to several GB. For better performance, you should care of
- avoiding concurrent updates (probably not an issue for a personal DB)
- creating relevant indices (they are essential if you are going to do a lot of analysis, running joins)
- running multiple inserts in a single transaction (if you need to insert a lot)
45 columns and 63 tables are well within SQLite limits.
jetxee
2010-08-24 08:37:00
A:
That sounds entirely practical, but you may wish to consider what the best way of representing the data is. In general, a database works best when each table represents exactly one simple relation, whereas it's all too easy with Excel to jumble everything together.
Donal Fellows
2010-08-24 08:44:13