Correlation

Created at 20 Sep 2018, 23:41

Join Discord

DelFonseca

Joined 25.06.2017

Correlation
20 Sep 2018, 23:41

Good evening,

I am developing a correlation bot but im stuck in the correlation calculation.

Result indicates that its not a number.

Follow the code below and thank you for your cooperation.

        private void CoefCorrel()
        {
            int index = 0;
            SubSymbolGetSeries = MarketData.GetSeries(SubSymbol, TimeFrame);
            var SubSymbolGetSymbol = MarketData.GetSymbol(SubSymbol);
            int CountOfBars = 24;
            double Result = 0;

            double sumX = 0, sumY = 0, sumXY = 0, sumX2 = 0, sumY2 = 0;
            double x, y, x2, y2;

            int index2 = GetIndexByDate(SubSymbolGetSeries, SubSymbolGetSeries.OpenTime[index]);

            if (index2 == -1)
                return;

            for (int i = 0; i < CountOfBars; i++)
            {
                x = MarketSeries.Close[index - i];
                y = SubSymbolGetSeries.Close[index2 - i];

                x2 = x * x;
                y2 = y * y;
                sumX += x;
                sumY += y;
                sumXY += x * y;
                sumX2 += x2;
                sumY2 += y2;
            }

            Result = (CountOfBars * (sumXY) - sumX * sumY) / Math.Sqrt((CountOfBars * sumX2 - sumX * sumX) * (CountOfBars * sumY2 - sumY * sumY));

            ChartObjects.DrawText("LabelCorrel2", "\n\n\n\n\n\n\n\n\nAverageSubSymbol: " + Result, StaticPosition.TopLeft, Colors.GreenYellow);
        }

        private int GetIndexByDate(MarketSeries series, DateTime time)
        {
            for (int i = SubSymbolGetSeries.Close.Count - 1; i >= 0; i--)
            {
                if (time == SubSymbolGetSeries.OpenTime[i])
                    return i;
            }
            return -1;
        }

@DelFonseca

Replies

afhacker
21 Sep 2018, 07:16

        public static double GetCorrelation(DataSeries dataSeries, DataSeries otherDataSeries)
        {
            double[] values1 = new double[dataSeries.Count];
            double[] values2 = new double[dataSeries.Count];

            for (int i = 0; i < dataSeries.Count; i++)
            {
                values1[i] = dataSeries.Last(i);
                values2[i] = otherDataSeries.Last(i);
            }

            var avg1 = values1.Average();
            var avg2 = values2.Average();

            var sum = values1.Zip(values2, (x1, y1) => (x1 - avg1) * (y1 - avg2)).Sum();

            var sumSqr1 = values1.Sum(x => Math.Pow((x - avg1), 2.0));
            var sumSqr2 = values2.Sum(y => Math.Pow((y - avg2), 2.0));

            return Math.Round(sum / Math.Sqrt(sumSqr1 * sumSqr2), 2);
        }

This function returns the correlation between two data series.

@afhacker

afhacker
21 Sep 2018, 07:21

For double generic collections you can use this method:

        public static double Correlation(IEnumerable<double> x, IEnumerable<double> y)
        {
            double xSum = x.Sum();
            double ySum = y.Sum();

            double xSumSquared = Math.Pow(xSum, 2);
            double ySumSquared = Math.Pow(ySum, 2);

            double xSquaredSum = x.Select(value => Math.Pow(value, 2)).Sum();
            double ySquaredSum = y.Select(value => Math.Pow(value, 2)).Sum();

            double xAndyProductSum = x.Zip(y, (value1, value2) => value1 * value2).Sum();

            double n = x.Count();

            return ((n * xAndyProductSum) - (xSum * ySum)) / Math.Sqrt(((n * xSquaredSum) - xSumSquared) * ((n * ySquaredSum) - ySumSquared));
        }

@afhacker

DelFonseca
21 Sep 2018, 20:36

RE:

afhacker said:

For double generic collections you can use this method:

        public static double Correlation(IEnumerable<double> x, IEnumerable<double> y)
        {
            double xSum = x.Sum();
            double ySum = y.Sum();

            double xSumSquared = Math.Pow(xSum, 2);
            double ySumSquared = Math.Pow(ySum, 2);

            double xSquaredSum = x.Select(value => Math.Pow(value, 2)).Sum();
            double ySquaredSum = y.Select(value => Math.Pow(value, 2)).Sum();

            double xAndyProductSum = x.Zip(y, (value1, value2) => value1 * value2).Sum();

            double n = x.Count();

            return ((n * xAndyProductSum) - (xSum * ySum)) / Math.Sqrt(((n * xSquaredSum) - xSumSquared) * ((n * ySquaredSum) - ySumSquared));
        }

Thank you very much for your feedback.
Please tell me how I can make correlation value print.

@DelFonseca

cTrader Ltd offers through its group of companies the cTrader platform. The information on this website is for general informational purposes only and does not constitute financial or investment advice. cTrader does not solicit retail investors. Reliance on this information is at your own risk.