Strategy using Neuro-Lab with ZigZag Output function

pedro1977

#1

10/14/2013 8:32 AM

Hi there, just wanted to share this strategy I saw in a youtube video in russian, using WL 6.2. I don´t speak russian, but I copied the code and I believe maybe the only thing needed to make it run is to update the ZigZag (ZIndicators) which apparently doesn´t exist anymore (Only errors I get, besides 4 lines of indicators). Maybe someone can help me update this to the new 6.5 standards (peak and through I suppose) ? If it´s too much trouble, please just point me in the right direction. Thank you.

Link to the video : https://www.youtube.com/watch?v=l0barFZdBmE
WL Screens go from 01:06:00 to 01:35:00 approx.

CODE:
Please log in to see this code.

Size:

Color:

Hi there, just wanted to share this strategy I saw in a youtube video in russian, using WL 6.2. I don´t speak russian, but I copied the code and I believe maybe the only thing needed to make it run is to update the ZigZag (ZIndicators) which apparently doesn´t exist anymore (Only errors I get, besides 4 lines of indicators). Maybe someone can help me update this to the new 6.5 standards (peak and through I suppose) ? If it´s too much trouble, please just point me in the right direction. Thank you.

Link to the video : [link]https://www.youtube.com/watch?v=l0barFZdBmE[/link]
WL Screens go from 01:06:00 to 01:35:00 approx.

[code]
// Input Script

using System;
using System.Collections.Generic;
using System.Text;
using System.Drawing;
using WealthLab;
using WealthLab.Indicators;
using NL;

namespace WealthLab.Strategies
{
	public class MyInputScript : WealthScript
	{
		public NeuroLab neuroLab = new NeuroLab();
                               
		protected override void Execute()
		{
	  Bars Bars = GetExternalSymbol("INDV13", "Index", true);
			DataSeries vol = WealthLab.Indicators.StdDev.Series(Bars, Close, 50, WealthLab.Indicators.StdDevCalculation.Sample);
			DataSeries fastma = WealthLab.Indicators.SMA.Series(Bars, Close, 10);
			DataSeries slowma = WealthLab.Indicators.SMA.Series(Bars, Close, 50);
			DataSeries RSI = WealthLab.Indicators.SMA.Series(Bars, Close, 50);
			neuroLab.Input(vol);
			neuroLab.Input((fastma - slowma)/slowma);
			neuroLab.Input(RSI);
		}
	}
}

[/code]
 [code]

// Output Script
using System;
using System.Collections.Generic;
using System.Text;
using System.Drawing;
using WealthLab.Indicators;
using NL;
using ZIndicators;

namespace WealthLab.Strategies
{
	public class MyOutputScript : WealthScript
   	{
		public NeuroLab neuroLab = new NeuroLab();

protected override void Execute()
		{
			ZigZag z = ZigZag.Series(Bars, 0.5);
			DataSeries Signal = new DataSeries(Bars, "Trade Signal");
			Signal [0] = 0;
			int i = 0;
			for (i = 1; 1 < Bars.Count; i++)
			{
				if (z.Bin[i]==0) Signal[i] = Signal[i] - 1;
				else Signal[i] = z.Bin[i];
			}

neuroLab.Output(Signal);
		}
	}
}
[/code]

[code]

// Strategy code
using System;
using System.Collections.Generic;
using System.Text;
using System.Drawing;
using WealthLab;
using WealthLab.Indicators;
using NL;
using ZIndicators;

namespace WealthLab.Strategies
{
	public class MyStrategy : WealthScript
	{
		protected override void Execute()
		{
			
			SetShareSize(1);
			double delta = 10;
			double mean = 50;
			PrintDebug("Start calc nn");
			Bars bb = new Bars("ss", BarsScale.Minute, 1);
			
			double scale = 100;
			if (Bars.Symbol == "INDV13") scale = 1;
			
			bb.Append(Bars);
			for (int b = 0; b < Bars.Count; b++)
			{
			
				bb.Close[b] = bb.Close[b] /scale;
				bb.Open[b] = bb.Open[b] /scale;
				bb.Low[b] = bb.Low[b] /scale;
				bb.High[b] = bb.High[b] /scale;
				
			}
			PrintDebug("Scale: " + scale);
			System.Diagnostics.Stopwatch wch = System.Diagnostics.Stopwatch.StartNew();
			DataSeries zz = NNIndicator.Series(bb, "SampleNet");
			
			PrintDebug("Neuro Net: " + wch.ElapsedMilliseconds);
			
			ZigZag zigzag = ZigZag.Series(Bars, 0.5);
			DataSeries Signal = new DataSeries(Bars, "Trade Signal");
			Signal [0] = 0;
			int i = 0;
			for (i= 1; i < Bars.Count; i++)
			{
				if (zigzag.Bin[i] == 0) Signal [i] = Signal [i-1];
				else Signal[i] = (zigzag.Bin[i]+1) *50;
			}
			for (i = Bars.Count - 1; zigzag[i] == 0; i--)
			{}
			for (int j = i; j < Bars.Count; j++)
			{
				zigzag[j]=zigzag[i];
				
			}
			
			PlotSeries(PricePane, zigzag, Color.Black, WealthLab.LineStyle.Dashed, 2);
			PrintDebug("Stop calc nn");
			ChartPane nnpane = CreatePane(50, true, true);
			PlotSeries(nnpane, zz, Color.Blue, WealthLab.LineStyle.Solid, 1);
			PlotSeries(nnpane, Signal, Color.Red, WealthLab.LineStyle.Dashed, 2);
			for (int bar = 0; bar < Bars.Count; bar++)
			{
				//  PrintDebug(bar);
				if (zz[bar] > mean + delta)
					// Sell signal
					if (IsLastPositionActive)
					{
						if (LastActivePosition.PositionType == PositionType.Long)
						{
							SellAtMarket(bar + 1, LastPosition, "NN Short");
							ShortAtMarket(bar + 1, "NN Short");
							
						}
					}
					else
					{
						ShortAtMarket(bar + 1, "NN Short");
					}
							
				
		
				else if (zz[bar] < mean - delta)
				{
				
					// buy signal
					if (IsLastPositionActive)
					{
						if (LastActivePosition.PositionType == PositionType.Short)
						{
							CoverAtMarket(bar + 1, LastActivePosition, "NN Long: Close Short");
							BuyAtMarket(bar + 1, "NN Long: open long");
						}
					}		
					else
					{	
						BuyAtMarket (bar + 1, "NN Long: open long");
					}
				}
				
			}
		}
	}
}
[/code]

Eugene

#2

10/14/2013 8:41 AM

ZIndicators must be a library developed by Russian company Zerich Investments a couple years ago. Unfortunately, they're no longer in this (Wealth-Lab) business. To try to obtain the library, you might want to contact their successors at:

http://www.wealth-lab.net/forum.aspx

As an alternative, check out this ZigZag class:

ZigZag Class

pedro1977

#3

10/14/2013 8:57 AM

Yes Eugene, you´re right, it´s from Zerich :) I already tried contacting them, let´s see. I have a ZigZag Strategy that I like, can I use it instead in Neuro-Lab ? How would I do that, please ? Do we have something similar developed ? Thank you.

Eugene

#4

10/14/2013 9:20 AM

I'm not familiar with this, but you might find it useful:

ZigZag indicators with NeuroLab

pedro1977

#5

10/14/2013 10:05 AM

Eugene, could you please just help me with this piece of code ? Using the ZigZag Class, but I get an error CS1061 : it doesn´t contain a definition for "Bin". I have no idea how to change that. Thanks.

CODE:
Please log in to see this code.

Eugene

#6

10/14/2013 10:10 AM

Look, you're confusing the Zerich ZigZag class with our ZigZag where there's no "Bin" property. Sorry, we can not help you with the code that is not supported by its creators. Don't merge different code mechanically, it won't work.

pedro1977

#7

10/14/2013 10:27 AM

I´m trying to use the class you showed me. Not trying to mix them. But since there´s no "Bin" property, I was just thinking if there´s a way to do what it´s doing using what´s available in WL libraries. It´s creating a binary Signal variable, right ? I know that´s what shows on his screen. I tried searching for binary in all the documentation and Wiki and found nothing, that´s why I´m asking for help. It´s my first week using WL and I really really like it, sorry for all the newbie questions. Just trying to learn this with no programming experience but very willing to put in the hard work. Is this very basic required knowledge from C# ? No problem, just let me know and I´ll go learn it.