Support Board
Date/Time: Sat, 23 Nov 2024 23:54:39 +0000
Post From: Offering To The Community: Custom Stock Index Creator
| [2016-12-13 08:36:41] |
| bjohnson777 (Brett Johnson) - Posts: 284 |
|
I'll start by apologizing... I missed an ACSIL function that's been added since I last read the doc page. This is the first thing I would need: sc.GetStudyNameFromChart() http://www.sierrachart.com/index.php?page=doc/ACSIL_Members_Functions.html#scGetStudyNameFromChart There's another missing piece I'll explain in a moment. The reason for needing sc.GetStudyNameFromChart() is that internally all studies are referenced by an ID number or index number. Any minor changes to the chart can change the ID number to something else (mentioning "EXACT location" in my previous post). That would mean that ACSIL would be accessing the wrong study and getting bad data. By getting the study name, it doesn't matter if the ID number changes. A test can be run against the name to first verify that it is correct. If not, the program can search the name list to see if it was moved or deleted from the chart. This provides a margin of safety. The other missing piece is being able to automatically load a study (VWAP in this case) to a newly opened chart by ACSIL. I don't see any way of doing this in the documentation. To be honest, to have ACSIL try and figure out what a new study needs is insanely complicated to program... but I had an idea around that yesterday. The workaround is to have the human create a Study Collection and get all the studies tweaked to desired values. After that, ACSIL would (in theory) just apply the Study Collection to the newly opened chart. This works well for automation and guarantees that a chart doesn't get missed (like the 100 charts out of the NASDAQ 100). For someone to individually add a study to 100 charts would probably take between 2 and 4 hours. For ACSIL to do it automatically would probably take about 5 minutes. I think I'm going to open a support request for this feature in a few minutes. This should be easy for the SC dev's to implement since most of the work has already been done in the deeper internals of SC. I don't know how long it will take them to implement this feature, though. Once this gets implemented, it shouldn't be too hard to create a full length chart (even on multiple time scales) for the number of stocks above/below VWAP. There might be some interesting chart patterns from this. The biggest thing slowing me down right now is my health issues. I have some vicious variant of Central Sleep Apnea where my brain forgets to breathe when I'm unconscious. I spend around 12 hours in bed hooked up to an artificial ventilator. If I'm lucky, I'm unconscious for most of that time, but my equivalent sleep compared to a normal person will be 3-4 hours per night. It's really tearing me up, and for the past several months it has been getting worse. For my next medical project/wetware hack, I'm going to use the OpenEEG project and create a brainwave interface to the computer. I'm hoping to use electrical signals to try and force my brain into deeper levels of sleep and stop the insomnia that's been plaguing me for the past several months. I've already done simpler forms of this and know it works, but I need something that is much stronger. The past 2 nights have been bad, and I'm pretty much useless today. If the SC dev's can get the missing function programmed, it will probably be 2 months before I can sit down and program this (that might be how long the SC dev's take to get their part implemented, anyways). I'd like to expand the code so my new program's framework can be used more generically as a stock scanner (kinda what I mentioned above). In a study format, it will still be a bit clumsy to use, but it should be far easier than the spreadsheet. And speaking of spreadsheets, I read somewhere in the docs that an external spreadsheet program could be used. Since I know C++, I've never bothered learning spreadsheets, but do you think an external spreadsheet could get around some of your current limitiations? It might be worth digging forther in the docs for. |