tag:blogger.com,1999:blog-6950062.post115629971588068447..comments2023-11-12T03:20:26.385-06:00Comments on TaylorTree: Development 0.1Mike Taylorhttp://www.blogger.com/profile/03074565631668167250noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-6950062.post-1156477257485725652006-08-24T22:40:00.000-05:002006-08-24T22:40:00.000-05:00epicjohn,The goal of the project is to take a set ...epicjohn,<BR/><BR/>The goal of the project is to take a set of trades from a system and apply dynamic allocation methods against those trades on a weekly basis. I need the ability to determine the current volatility of a stock within the portfolio and compare against the overall volatility limit of the portfolio. If too high...adjust the position (shares) down until the risk is back to acceptable levels.<BR/><BR/>Of course, I'd also like to test the ability to adjust upward if volatility drops. Many people say adjusting upwards shouldn't be done...but I want to test it regardless.<BR/><BR/>So, I'll need the capability to read in a set of trades, determine risk of the positions while being held, and cut back or add to as necessary until the system's trailing exit kicks in.<BR/><BR/>But, if I'm going to go that far with the development...I might as well code the entry and exit portion to output those trades instead of my current method of using Wealth-lab and outputting the results to csv files.<BR/><BR/>Thanks for the R/python link...that is right up my alley. I use R a lot in examining the system csv files. So, that could prove very helpful in the project. I'll take a look at matplotlib as well. Though, my plans are to keep this a batch-oriented environment.<BR/><BR/>If you try the pytables out...let me know what you think about it. I hope to try it out a bit this weekend. Maybe write a post comparing the create, insert, fetch, drop of sql versus the pytable equivalent.<BR/><BR/>Have a great weekend!<BR/><BR/>MTMike Taylorhttps://www.blogger.com/profile/03074565631668167250noreply@blogger.comtag:blogger.com,1999:blog-6950062.post-1156458444535525442006-08-24T17:27:00.000-05:002006-08-24T17:27:00.000-05:00oh by the way, check out http://rpy.sourceforge.ne...oh by the way, check out <BR/>http://rpy.sourceforge.net/<BR/><BR/>makes it easy to use R from inside python and it meshes well with Numeric arrays and all that.<BR/><BR/>Another very useful package is matplotlib. It even has some usable stock charting examples for candlesticks.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-6950062.post-1156458166828106182006-08-24T17:22:00.000-05:002006-08-24T17:22:00.000-05:00What is the goal of your project exactly? I'm in a...What is the goal of your project exactly? I'm in a similar situation to you and have several half-finished market related projects in python/postgres sql, but i think i should maybe switch to pytables. since i really dont need a relational db for time series data. doing stuff like getting todays open and yesterday's close for all days is harder in sql without some extra trickery.Anonymousnoreply@blogger.com