Noise measuring functions

Hi Eddy,

The only python library I was able to find that supported one of these methods is the FractalDimension library. However, all of the methods have the formulas defined in Kaufman's book. I recommend just creating a custom indicator for each of the formulas. This thread demonstrates how to create a custom python indicator.

Best,
Derek Melchin

Thanks for the tips Derek I'll check that library out and will look into creating the formulas in an indicator. Can you recommend any resources I can look at to learn how to read the formulas? 

Thanks

Hi Eddy,

To read the formulas in Kaufman's book, we need to understand: sigma notation, absolute value, derivatives, and logs. Google has countless resources on these topics. Don't hesitate to ask the community either!

Best,
Derek Melchin

Thanks Derek, that's really helpful, I will get cracking researching those topics!

Eddy Bennett  : were you able to make any progress with noise measuring functions? I'm looking to measure Efficiency Ratio, so if you've already coded anything for this, it'd be great if you could share!

Hey ekz, I modified the Kaufman moving average indicator to get the efficiency. I haven't tested it thoroughly but you might find it useful. 

using QuantConnect.Indicators;
using System;
using System.Collections.Generic;
using System.Linq;
using System.Text;
using System.Threading.Tasks;

namespace KaufmanEfficiency
{
    class KEfficiency : WindowIndicator<IndicatorDataPoint>, IIndicatorWarmUpPeriodProvider
    {
        private decimal _sumRoc1;
        private decimal _periodRoc;
        private decimal _trailingValue;

        public KEfficiency(string name, int period) : base(name, period + 1) { }

        public int WarmUpPeriod => Period;

        protected override decimal ComputeNextValue(IReadOnlyWindow<IndicatorDataPoint> window, IndicatorDataPoint input)
        {
            if (Samples < Period)
            {
                if (Samples > 1)
                {
                    _sumRoc1 += Math.Abs(input.Value - window[1].Value);
                }
                return input.Value;
            }

            if (Samples == Period)
            {
                _sumRoc1 += Math.Abs(input.Value - window[1].Value);

                // At this point sumROC1 represents the 
                // summation of the 1-day price difference
                // over the (optInTimePeriod-1)
            }

            var newTrailingValue = window[Period - 1];
            _periodRoc = input.Value - newTrailingValue.Value;

            if (Samples > Period)
            {
                // Adjust sumROC1:
                // - Remove trailing ROC1 
                // - Add new ROC1
                _sumRoc1 -= Math.Abs(_trailingValue - newTrailingValue.Value);
                _sumRoc1 += Math.Abs(input.Value - window[1].Value);
            }

            // Save the trailing value.
            _trailingValue = newTrailingValue.Value;

            // Calculate the efficiency ratio
            return  (_sumRoc1 <= _periodRoc) || _sumRoc1 == 0 ? 1m : Math.Abs(_periodRoc / _sumRoc1);
        }

        public override bool IsReady => Samples >= Period;

        public override void Reset()
        {
            _sumRoc1 = 0;
            _periodRoc = 0;
            _trailingValue = 0;
            base.Reset();
        }
    }
}

Thanks @sean, I did see that lean's KAMA indicator calculated ER, but couldn't use the directly since I am working in python.

I'm going to go ahead and write the indicator from scratch in Python. Will share here when done. 

.ekz., Thank you in advance for your contributions.