Page 1 of 1

Advance Bars Desired (for historical testing purposes)

Posted: Nov 17 2006
by denizen2
It would be VERY useful to be able to reference price bars in the 'future', for historical testing purposes, e.g., as price[-N].

Or perhaps it would be easier to implement without changing the compiler by just adding a 'special' function for advanced bars for only the prices, Open, High, Low, Close, and maybe name them, _Open(N), _High(N), _Low(N), _Close(N), where the value of N could be a negative or positive number.

I assume it would be appropriate to then use positive N for bars in the past, and negative N for bars in the future, while any bars beyond the end of the data-series would just be assigned, say, the same value as the end-value.

Posted: Nov 17 2006
by Alex Kramer
Thanks for the suggestion, we'll consider this

Posted: Nov 22 2006
by damageboy
Hi denizin2,
You could install ELCollections + ADE to achive what you desire.

Using ADE you can basically reference any bar from any timeframe that you have pre-recorded in advance, so for all purposes, you could reference a year in advance assuming the ADE data exists for this bar.

Posted: Nov 22 2006
by denizen2
Hi denizin2,
You could install ELCollections + ADE to achive what you desire.

Using ADE you can basically reference any bar from any timeframe that you have pre-recorded in advance, so for all purposes, you could reference a year in advance assuming the ADE data exists for this bar.
Yes, or maybe, it will do the job... but at the expense of a significant learning curve and/or increased CPU performance hits. :?:

My 'desired' solution is to be able to access 'future' data with very little learning effort and little pre-planning, other than merely calling a pre-built function that specifically returns the Open, High, Low, and Close, for ANY kind of bars (intraday-seconds, minutes, or N-Ticks, range-bars, or daily, etc), and without the extra overhead of text-files being generated by ADE+ELCollections approach.

Your suggestion is welcomed and probably a valid 'solution' under some conditions. It seems to require too many support functions to learn and use, if all you want to do is just reference N-bars ahead for *any* price-data. I also don't want to be bothered with having to setup and save a particular price-data series, just in order to reference future bars for a new idea that I want to test.

Am I missing something? Is it more simple than I have concluded? Are there any 'restrictions' on what kind of bars you can 'access' with the "ADE+ElCollections" approach?

Posted: Nov 23 2006
by Pipscalper
The possibilty to use bars in the future should also be implemented in Pleditor and the drawings like the Gann square which cannot be placed beyond the chart, now we have to set the sufficient space to the right.

Posted: Nov 23 2006
by Alex Kramer
Please tell us more about how you envision such projected bars, how are they to be created?

Posted: Nov 24 2006
by Pipscalper
For instance if i would like to project a time and price into the future,based on previous swings, Pleditor will not except the future dates and it cannot be plotted on the chart without setting sufficient space to the right,the move right arrow in the toolbar cannot move beyond the space settings.
the same goes for the gannbox,you cannot move it to the right if spacesetting is not sufficient.

Posted: Nov 24 2006
by Alex Kramer
Why not assign a lot of space to the right in advance, if you're going to use Gann's squares on this chart?

Posted: Nov 24 2006
by Grailtrader
I am not currently a user of MultiCharts...but am interested in it.

When adding more space to the right to plot a Gann Square the chart will not then scroll normally as bars are updated, (At least this is the case in TS.) and one should have to not mess around with scrolling a chart manually after say plotting a very long Gann Square into the future, and then continually manually update scrolling as the chart updates. Ideally one would want to leave the standard number spaces to the right that one may have set on each chart window when plotting something like this. Also in TS at present...it is almost impossible to determine the end date and time of a trendline (for developing studies such as a Gann Square) into the future when one only knows the number of desired bars in length for a trendline...this would be a great feature to have in Multicharts.

Thank You....

Posted: Nov 27 2006
by Alex Kramer
Thanks for the interest, but could you please provide more visual examples of your problems, with screenshots maybe? We're doing guesswork from the descriptions you've provided - for instance, I tried applying a Gann square on a ticks symbol in MultiCharts and watched the chart line pass through the square and continue on - I'm not sure if this is what in your words does not work in TS.

Posted: Nov 28 2006
by Pipscalper
Alex draw a Gann box on the right side of a chart and then clone the box and shift it to the right, the box cannot be draw beyond the right side of the chart. So we have to add spaces to the right,which is not a big problem,but whith larger gann boxes the details on your chart get to much compressed. If you apply these boxes on a 1min chart it gets annoying to constantly change all the settings.