LukeB

This may be what you are looking for - to have the boxes grow in the II_PipAccumulator indicator:

Find the code where the existing two "ObjectCreate("Box_" + ctr, OBJ_RECTANGLE"..."'s are. Insert the below code just after those two if blocks.

if ( (BuyActive) && (i==1) )

{

ObjectCreate("Box_" + ctr, OBJ_RECTANGLE, 0, t1,HH,Time[0],LL); // Probably quicker to recreate than test.

ObjectSet("Box_" + ctr, OBJPROP_STYLE, STYLE_SOLID);

ObjectSet("Box_" + ctr, OBJPROP_COLOR, up_box_color);

ObjectSet("Box_" + ctr, OBJPROP_BACK, True);

}

if ( (SellActive) && (i==1) )

{

ObjectCreate("Box_" + ctr, OBJ_RECTANGLE, 0, t2,LL,Time[0],HH);

ObjectSet("Box_" + ctr, OBJPROP_STYLE, STYLE_SOLID);

ObjectSet("Box_" + ctr, OBJPROP_COLOR, dn_box_color);

ObjectSet("Box_" + ctr, OBJPROP_BACK, True);

}

Some have wanted to have the boxes grow in the II_PipAccumulator indicator. To make them do so:

Find the code where the existing two "ObjectCreate("Box_" + ctr, OBJ_RECTANGLE"..."'s are. Insert the below code just after those two if blocks.

if ( (BuyActive) && (i==1) )

{

ObjectCreate("Box_" + ctr, OBJ_RECTANGLE, 0, t1,HH,Time[0],LL); // Probably quicker to recreate than test.

ObjectSet("Box_" + ctr, OBJPROP_STYLE, STYLE_SOLID);

ObjectSet("Box_" + ctr, OBJPROP_COLOR, up_box_color);

ObjectSet("Box_" + ctr, OBJPROP_BACK, True);

}

if ( (SellActive) && (i==1) )

{

ObjectCreate("Box_" + ctr, OBJ_RECTANGLE, 0, t2,LL,Time[0],HH);

ObjectSet("Box_" + ctr, OBJPROP_STYLE, STYLE_SOLID);

ObjectSet("Box_" + ctr, OBJPROP_COLOR, dn_box_color);

ObjectSet("Box_" + ctr, OBJPROP_BACK, True);

}

Not sure which ST indy everyone is using, but this one is MTF Non repainting Supertrend

 

Probably better to use this one if including into an EA.

 

http://www.multiupload.com/5EVYRKT8GN

 

I'm not sure the Supertrend indicator here is good, or at least, it seems to do a lot of extra work.

I believe it loops through the entire history of bars with every tick. Also, I'd like to get comments or explanation of what is going on in PlotPoint and CleanPoint.

Anyway, here it is re-written to display in its own window, and to show the four buffer values for the last four bars.

 

http://www.multiupload.com/IEPZJO0YAR

 

Any help cleaning it up / explaining what is going on appreciated - or if there are references to versions that may be better.

Fortuna - updated your changes.

http://www.multiupload.com/IFL8HUQB8E

 

Added GetNearCandle() method.

There is no perfect way to get and calculate historical probabilities and have them match a live candle probability calculation. I think the nearest candle method is the best compromise. The code now has three ways to handle getting the candles for the indicator - I defaulted it to the Nearest Candle method:

II_PipAccumulator method:

- Advantage: Current probability is responsive and uses current signal data.

- Disadvantage: Historical candle probability calculations can use data far removed from the time of the candle being addressed.

Last Candle Method:

- Advantage: Current and historical candles are calculated the same way.

- Disadvantage: Current Probabilities are calculated with stale data, making them unresponsive to the current price action.

Nearest Candle Method:

- Advantage: The current probability calculation is responsive and uses current signal data. Historical Candle calculations use data from other timeframes 'near' to what would have occurred in a live candle calculation.

- Disadvantage: Historical Candle probability calculations are not exactly the same as the Current / Live candle calculation.

Fortuna - "the difference between Last Candle and II Pip Accu method"

 

The Last Candle method is supposed to find the last candle completed for each Period for the candle the probability is being calculated for (the Time[1] candle for all Periods at time the current Period() candle became the Time[1] candle), versus using the same candle number for all Period's.

 

Example - Assume you are getting the probabilities for the M5 Time[12] candle and that the Time[0] candle has just started (and no time has passed since it started, as our resolution is to the second with datetime variables). The Time[12] M5 candle ended 55 minutes ago ( (5*12)-5 = 55, started 60 minutes ago and ended 55 minutes ago). Unless I'm mistaken, the II_PipAccumulator will calculate that Time[12] M5 candle's probability using the Time[12] M1 (minute) candle (ended 11 minutes ago, and not in existence at the time that Time[12] M5 candle was created) and do similarly for other Period calculations. To be always consistent, the Time[12] M5 candle's probability should be calculated with the Time[56] M1 candle (started 56 minutes ago and ended 55 minutes ago ), as that is the last complete M1 candle in its Period.

 

Now, I think the logic in that candle selection function needs to be reviewed to see if it really does the job. A task for another day.

I'm embarrassed - as soon as I tested on other platforms (Lucror) it failed, with WindowNum = -1. So, most of you who downloaded must be looking at nothing. Hey, it worked on the IBFX Australia platform. I deleted the other uploads as unfit for human consumption.

 

Here it is, fixed, Works on IBFX Australia, Lucror - but my Oanda Demo account does not have enough MN1 bars of history. I changed MinBarCount to 50 and it works on Oanda, but the numbers are different.

 

http://www.multiupload.com/IFL8HUQB8E (Signal_Amalgamate.mq4)

 

I changed the name to Signal Amalgamate - I don't think it's much like II_PipAccumulator.

I think I'm done with this unless I get feedback to do things that tickle my fancy.

http://www.multiupload.com/YR15TR6YI6

- updated to limit history, per fortuna.

- updated bug, processCandle --; changed to processCandle ++; (increase the candle count to go back in time).

- added an object to display all the timeframe probabilities.

Fortuna - this just provides a history display of what the probabilities were - it was created so I could see what the probabilities looked like over time. It is not a trading tool - but having the indicator separated from the trading tool and able to be looked at over time could assist in analysis to make improvements for inclusion in the trading tool.

Seems like the technical side of this discussion has stopped with soundFX's contributions. I created this, maybe it will generate some interest there.

http://www.multiupload.com/IFL8HUQB8E

I wanted to see what the Master Probability looked like over time, so I made it (and all TF Probability) a separate window indicator, with everything else removed.

It was a weekend coding effort, without a lot of testing, but it produces the same ending probability as the II_PipAccumulator (which it is mostly a re-organization of), so I think its correctly coded. I was somewhat surprised at it's movement - I thought it would be smoother and more regular. I was surprised at how little the buy and sell indications were related to the Master probability.

I introduced a function "GetProcessCandle" that I think is more accurate (at least in thought), but produced a much smaller effect than I expected. I tried to improve readability of the Master Probability calculation and make it a little more efficient. soundFX would be the guy to look it over and see if it should be implemented this way in II_PipAccumulator.

It's also called Pips Chaos. I'm pretty sure its the exact same thing - and there is already a thread.

Re: [Req] Forex Income Engine - EA

 

Cyrillic, "what did you want to get out of it?" Nothing much - just a technique to follow that would lead to profits without scorching account pull downs. Is that the holly grail? I guess if it were easy, there would be no opportunity.

Re: [Req] Forex Income Engine - EA

 

Rio, Cyrillic, Did you code FIE 1 or FIE 2? I purchased FIE 1 and have tested quite a bit - I believe that just following its rules will not be long term profitable. Hence, no purchase of FIE 2. It would be nice to get more out of this than just a pretty presentation though.

I wrote this ea some time ago, only coded in buy logic to see if it was profitable and it failed... So not worth to effort to make the sell logic..