Thanks Roy, i'll follow that up too may even be my new laptop-- just gone back to be fixed ( again !!) I noticed that on one exploration, the long Ma was above?? the short MA in an uptrend instead of the opposite -- checked it twice but still can't believe it
I'll test it on this machine too thanks Keith ----- Original Message ----- From: Roy Larsen To: [email protected] Sent: Sunday, February 02, 2014 7:29 AM Subject: Re: [EquisMetaStock Group] accuracy of code signals Hi Keith I wonder if you're scaling the MAs to the same scale as prices on your charts. It seems to me that it could be the charts that appear to be wrong rather than the exploration. Another possibility is that you have an historical date set for the exploration (Edit exploration and click Options) rather than the current date. This would give results relevant to date in question rather than the current date. As Bill has said, the number of bars loaded can affect results, but if too few are loaded this would usually lead to N/A results. If Minimum is the setting you're using then MetaStock will probably be loading about 30% more than the number actually required. It's when you're using Exponential MAs that you'll get erroneous results with Minimum bars. Regards Roy ----- Original Message ----- From: Bill Hebert To: [email protected] Sent: Sunday, February 02, 2014 7:23 AM Subject: Re: [EquisMetaStock Group] accuracy of code signals Four things to check or truy, 1) One thing to check and seeis how much data you are loading to make sure you have enough data for the indicators to work properly 2) Based on what you have written - (high of today to be greater than a 20MA) AND (the 20MA to be above a 40MA) Try ColA = (Mov(c,20,S) >=Mov(C,40,S)) AND (H >=Mov(C,20,S)) and leave off the Volume Filter for now to see of that is right before adding the Volume Filter or you can try ColC = ColA AND ColB but leave off the Volume Filter until you get it working without it first 3) Try doing something very simple and basic and get that to work ad then add things to it and check each addition as you add it. 4) Make sure you are doing a Daily Exploration and not a Weekly Exploration ----- Original Message ----- From: a To: [email protected] Sent: Friday, January 31, 2014 7:22 PM Subject: [EquisMetaStock Group] accuracy of code signals I'm just testing to get a reliable signal from MS 7.2 but little luck so far in a simple exploration, I asked for the high of today to be greater than a 20MA and the 20MA to be above a 40MA so ColA Mov(c,20,S) >=Mov(C,40,S) ColB H >=Mov(C,20,S) with a volume filter out of 2000 stocks , I got about 12% & most were wrong so I changed ColA to Cross(H,Mov(C,20,S) & this time I got about 1.5% -- pretty good I thought but when I looked at the charts, all showed the 20ma above the 40ma , and all had the high above the 20ma, but none crossed today ! -- most crossed 5-10 days ago I thought the Cross would be tight but not true am I missing something here? I'd appreciate some advice as to how to sweettalk MS into doing what I asked it to thansk Keith __________ Information from ESET NOD32 Antivirus, version of virus signature database 9368 (20140201) __________ The message was checked by ESET NOD32 Antivirus. http://www.eset.com __________ Information from ESET NOD32 Antivirus, version of virus signature database 9369 (20140201) __________ The message was checked by ESET NOD32 Antivirus. http://www.eset.com
