[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Advances in TA: Event flags



PureBytes Links

Trading Reference Links

There is an effective work around to do this by making an artificial 
ticker for each stock and then using the Foreign() call for testing.

For example, for IBM create a ticker IBM_Flags
Put your info into the various price and volume field. That will 
give you 6 data items. If you want 10 fields, divide the O, H, L, C 
fields in half using the decimal point. It is easy to use AB's 
Fraction and Integer calls to get what you want.

If you need a lot of data fields, you can "pack" the data into each 
digit of the price and volume fields. And then use AB's Fraction and 
Integer calls with appropriate division and multiplication to get 
the data items separated. Theoretically you should be able to get 6 
digits in the price fields and 9 in the volume and OI fields. But 
for safety you can use 5 from each. Also, you can pack 9 digits into 
the volume and oi fields, but you can only extract 6 digits since AB 
usings changes the data in these fields to decimal data (only 6 
significant digits) to do the multiplication and division functions. 
Even with just using 5 in each of the 6 regular fields, that gives 
you 30 data fields, each with a range of 0 to 9.

b

--- In amibroker@xxxx, 520001085531-001@xxxx wrote:
> Some thoughts from a new user:
> 
> What is missing in today's EOD TA packages is the
> lack of storing "events" with the daily data and the ability
> to use this information in formulae. 
> 
> These events do not need to take up much space.
> In fact they can be bit fields (Yes/No, True/False values).
> Using only 4 bytes for this would allow us such 32
> independent event flags.
> 
> These event flags could be picked from a predefined
> list, or be user-defined. For example:
> 
> COMPANY_CRISIS
> MARKET_CRISIS
> BAD_NEWS
> GOOD_NEWS
> BAD_RATINGS
> GOOD_RATINGS
> STOCK_SPLIT
> 
> ER_BEFORE_MARKET_OPEN
> ER_DURING_MARKET_OPEN
> ER_AFTER_MARKET_OPEN
> ER_BELOW_EXPECTATIONS
> ER_MEETS_EXPECTATIONS
> ER_ABOVE_EXPECTATIONS
> ...
> 
> Such events have an impact on the price. And they answer
> the question of (dramatic) price changes on a specific day(s).
> 
> By having this (optional) information with each EOD data
> one could develop better strategies/programs by filtering
> any of the (un)wanted events.
> 
> I hope, one day AmiBroker will have this feature implemented :-)
> (... or is it already possible? :-)