Recent content by John Tseng

Good points. There's really nothing explaining what's special about this system compared to others. I've added more details to the readme. I haven't done most of those things you're talking about. Mostly I just wanted to share with the community what I've done. I like your idea about...

=) Not only is the time slow, I had to write those two subqueries to tell SQL to do two nested loop joins. It looks ugly as well. (CTE was probably overkill, though.) I knew SQL was slower. Especially at open/close. I don't know how to get SQL to do this in one pass. GroupBy spools it up for...

Just for kicks, I wanted to try this out. Here's my sql for EURUSD: SET STATISTICS TIME ON DECLARE @ticks int = 100; with t as ( SELECT *, ROW_NUMBER() OVER (ORDER BY Ticks.Time ASC) TickIndex FROM dbo.Ticks WHERE Ticks.PairId = 188 AND Ticks.Time BETWEEN '2012-1-1' AND...

I've created my own forex back tester, and I've decided to share it. It's available at: https://github.com/johntseng/ForexRunner I realize it's pretty simple, but it's gotten me started. Any comments or suggestions are welcome.

Within a file, data is always stored in sequence. The file may be fragmented across different parts of the disk, but that's where defrag comes in. Where multiple files are stored is up to your filesystem. This is usually fairly random. I would have to go along with the other guys who favor...

We've gone through a few decades of optimizing disk access, so there's actually quite a bit of complexity here. 1) You are right that somewhere along the chain, disk accesses will be reordered to reduce overall seek time. The O/S is, however, not the best place to do it, since it doesn't know...

This is interesting. Most of the time network uptime is monitored from the outside world to see if your server is available, but you want to monitor it from the inside. For external availability, something like Pingdom is good. For monitoring from inside the server, New Relic's server...

I like your scheme. I'm sure it's incredibly fast. Since you know your application very well, you don't need to provide a general data interface that SQL provides. I think, though, it's not worth it at this stage for me to use such a scheme. Once I put in any non-trivial strategy into my...

I'm back after a few iterations. I now have two systems. The time based strategy that processes every minute is now a huge SQL script. It takes all my tick data, builds bars, computes intermediate data, and spits out trades for a bunch of different parameters all in one CTE (not quite as big as...

I've taken a different approach than blah. My database schema is extremely simple: Instruments table and Ticks table. I compute bars on the fly. So far it hasn't given me issues, but I'll probably create an indexed view if that's a problem. (Here's where I love Postgres: Materialized views (in...

This start small and iterate idea has become quite a large movement (Agile method). I think it plays well with "premature optimization is the root of all evil". I like your idea on TradeLink, though. It would be good to learn from what others have already done.

I'm doing a mix of both. I have ticks in the database, and my events are seconds. From what I've seen so far, other trading frameworks give you a stream of data, and it is the responsibility of your algorithm to store the data however you like (plus a few charting features like bars). I'm...

What are some resources you would recommend for a new day trader? I am a software developer looking to get into day trading. There's another thread asking for good books for investing, but I would like to know what kind of resources there are for day traders. Also, I am open to resources that...