Quanting News

Mike and I (mostly Mike) worked early this week on the testing platform, with good results. He got it setup so I can put all the data I want in the database (I'll eventually get about 5 years worth in there), choose the dates I want to test, run the test, choose the dates I want to chart, choose the data to chart, and view the chart. I can even zoom in and zoom out (though only to original dimensions for now). We still have some work to do, like getting it setup to read a file for algorithm variables so I don't have to hack the code everytime I want to tweak it, but it's in a state now that I can run tests quickly and look at the results pretty effectively.

Unfortunately, looking at the first few sets of tests, it looks like I may need to change the algorithm some. I'm getting good data on some parts of the algorithm, but there are certain other data points I am missing. I'm getting some decent results, but I don't know if that's just luck or if it's a decent algorithm that could be better. If I could get the other data points and then tweak the components a tiny bit, I could get some fantastic results. I'm just going to have to play with it to figure out how to get those other data points somehow without getting false positives. That last bit is going to be the tricky part.

So, things are progressing well, but I'm a little behind the schedule I had set for myself. I wanted to know if this thing was going to work by mid-February, before I went to Boston, but I don't think I'm going to make it at this point. The initial returns were good, about 1/3 to 1/2 of our target return, but again, that could just be luck. All-in-all though, things are progressing well and I should know something soon. Anyways, I'm going to go get some editing out of the way and maybe eat some soup. Yumm.