On Sunday 09 April 2006 07:30 pm, Tomasz Janeczko wrote: > "I'd like to see clear documentation on all this in the reference manual. " > > How about this: > http://www.amibroker.com/guide/h_understandafl.html
Yes, that is pretty good. In particular: "When AFL is evaluating statement like this ( High + Low )/2 it does not need to re-interpret this code for each bar. Instead it takes the High ARRAY and Low ARRAY and adds corresponding array elements in single stage." I remember reading that now, early on in the process of learning so many other things. So why did I get confused? I am not embarrassed to admit it. Why do many people get confused? Why does it drive some of them nuts? Mostly, for me, it was when I was trying to figure out how to set up realtime trading, for which there is only minimal help, I started making up theories about how I thought it ought to work. Dangerous idea. I still claim it would be much faster if AFL did in fact re-interpret the code of each new bar as it came in (it would have to store previously calculated bars of data), rather than to re-interpret for all the bars. For a thousand bars, it would be a thousand times faster. But that is a very different processing model. It would help if the docs for each of the functions either gave a complete definition, or referenced where to find more details. For example, the doc for Cum (http://www.amibroker.com/guide/afl/afl_view.php?id=35) gives a text definition: "Calculates a cumulative sum of the ARRAY from the first period in the chart.". That seems straightforward enough, but there are a few fuzzy points. What is the value for the first period? (== Array[0]) What is the "first" period when there is a date range in effect for scanning? What happens when a null value is encountered? It would be good to spell out on that page how Cum(array) differs from array + ref(array, -1), since some people who are confused about this question will find their way to this page in the docs. This might also be a good place to reference explicit loops, since one of the main reasons people might use a loop is to do something similar to the cum function. The docs would be adequate, I think, if there were many more cross references between relevant parts of the docs that already exist. It is more frustrating to find things without good cross references, and good indexing. I don't mind suggesting improvements to the docs, if you don't mind receiving them. I am mostly thinking of all the new people who will find their way to AmiBroker, given its growing success, and how to deal with an ever greater influx of frequently asked questions and frequent confusions, such as mine. dan Please note that this group is for discussion between users only. To get support from AmiBroker please send an e-mail directly to SUPPORT {at} amibroker.com For other support material please check also: http://www.amibroker.com/support.html Yahoo! Groups Links <*> To visit your group on the web, go to: http://groups.yahoo.com/group/amibroker/ <*> To unsubscribe from this group, send an email to: [EMAIL PROTECTED] <*> Your use of Yahoo! Groups is subject to: http://docs.yahoo.com/info/terms/
