ooples / ooplesfinance.stockindicators Goto Github PK

While testing I noticed a significant difference in the Keltner Channel calculation that is produced by this library when compared to what I see displayed in Trading View using what appears to be the same input and was hoping to get some clarity on why that may be.

An example of this is AAPL on 2023-01-27. The Keltner upper band in Trading View is 144.91. The Keltner Channel Upper Band produced by this library is 152.8937. The default inputs used by the CalculateKeltnerChannels method seems to the same as the input used in Trading View.

Please let me know if you need any more info.

Hi Franklin,

Thank you for this collection of indicators!
Here are some suggestions which you may decide to apply:

1. Design each indicator as a simple function
   This kind of design should be considered because of multiple reasons:

• Imagine having an infinite stream of data, this will allow to save memory and maybe even increase performance.
• Having simple functions allows composition
• Having simple functions allows easy testing and rewriting to different language.
  Please let me give you quick example of simple SMA in Java that covers all three points above:

public static class SMA implements Function<BigDecimal, BigDecimal> { 
   private final var buffer = new LinkedList<>()
   private final int length
   private var sum = BigDecimal.ZERO

   SMA(int length) {
     this.length = length
   }

   @Override
   BigDecimal apply(BigDecimal currentValue) {
        if (buffer.size() == length) {
            sum -= buffer.removeFirst()
         }
        sum += currentValue
        buffer.addLast(currentValue)
        // at the end calculated SMA should be rounded to SOURCE scale
        return roundValue(sum / buffer.size(), currentValue)
  }
}

So it is basically a function, that accepts decimal value and produces new one. It has some minimal internal state like a self buffer etc. Usage is quite simple, it can be applied now to any datasource, streaming or not. RSI from SMA will look like this:

var datasource = candlesticks.map(c -> c.close)
     .map(new SMA(3).then(new RSI(14)))

I think you may improve your code in this way if you agree.
2. Rounding of values is a problem now. You used 4 decimals by default in any calculation, but I think this is not correct.
Whenever you know the scale you should use it. For example any moving average should use scale of original sourced value. RSI should/may use just 2 decimals. SMA from RSI should have the same 2 decimals for calculated value.
3. I tried to use some cryptocurrency candlestick data and many indicators failed to generate any useful data, I even couldnt find simple signal for RSI with StrongSell/StrongBuy values, which is weird.

The most important part is point with number 1, I do not have much experience with C# to help you with that. Please let me know what do you think.

Request to add this very interesting volatility index, based on a fundamental principle of statistics.

The author is a seasoned pro trader and claims it beats most other trend indicators in backtesting.

Only requires half-a-dozen lines of straightforward code. Would do it myself, but new to both C# and Git so a pull request would be a bit of a learning curve. If you can manage to add this in the midst of your new parenthood it would be appreciated!

Explanation here, along with code in C:

https://financial-hacker.com/the-market-meanness-index/

Thanks for all your hard work producing this comprehensive library!

If you run the following code on any data:

stockData.CalculateRelativeStrengthIndex();

There will be an StackOverflowException. I tracked it down to the following lines:

    /// <summary>
    /// Adds the rounded.
    /// </summary>
    /// <param name="list">The list.</param>
    /// <param name="value">The value.</param>
    public static void AddRounded(this List<decimal> list, decimal value)
    {
        list.AddRounded(Math.Round(value, 4));
    }

This must be:

    /// <summary>
    /// Adds the rounded.
    /// </summary>
    /// <param name="list">The list.</param>
    /// <param name="value">The value.</param>
    public static void AddRounded(this List<decimal> list, decimal value)
    {
        list.Add(Math.Round(value, 4));
    }

OoplesFinance.StockIndicators.csproj contains references to a missing Images folder, reference to C:\Users\cheat\Downloads\Favicon.jpg and <PackageIcon>Favicon.jpg</PackageIcon>. After removing these I could get it to build. Overall StockIndicators.csproj could use some cleaning up.

This is really an impressive project BTW! Where do the indicator implementations come from? Did you write all 510 of them by hand?

Can't find any mention of the open source license anywhere.
Sorry if I missed it.

Hey,

love your lib, using long time, many times first :3

second, maybe you can add the UT Indicator Signals to your lib?

original here: https://www.tradingview.com/script/n8ss8BID-UT-Bot-Alerts/

many thanks :)

Hello!

I wanted to try out this awesome library. Testing the first indicator, Bollinger Bands like this i get the following error:

System.InvalidOperationException: 'Sequence contains no elements'

Stacktrace:

   at System.Linq.ThrowHelper.ThrowNoElementsException()
   at System.Linq.Enumerable.Average(IEnumerable`1 source)
   at OoplesFinance.StockIndicators.Calculations.CalculateSimpleMovingAverage(StockData stockData, Int32 length)
   at OoplesFinance.StockIndicators.Helpers.CalculationsHelper.GetMovingAverageList(StockData stockData, MovingAvgType movingAvgType, Int32 length, List`1 customValuesList)
   at OoplesFinance.StockIndicators.Calculations.CalculateBollingerBands(StockData stockData, MovingAvgType maType, Int32 length, Decimal stdDevMult)
   at OoplesFinance.StockIndicators.Calculations.CalculateBollingerBands(StockData stockData)
   at antihodl2core.LearnDataGenerator.SkenderMinutes.Run() in F:\src\antihodl2\antihodl2core\LearnDataGenerator\SkenderMinutes.cs:line 104
   at antihodl2core.Form1.<learningDataGenerator_Click>d__7.MoveNext() in F:\src\antihodl2\antihodl2core\Form1.cs:line 69

Not sure what could be missing here.

Hey,

You have a normal STC Indicator in your lib, maybe you can add this too? Its a modded one with macd added :)

https://www.tradingview.com/script/WhRRThMI-STC-Indicator-A-Better-MACD-SHK/

many thanks :)

Hi.
Thank you for this great library.

I'm trying to use your library in my project and I installed using nuget.

I'm trying to calculate the Supertrend indicator but it's returning all same values - all items in the OutputList contains same values)
Can you please provide sample code to calculate the Supertrend indicator?

Here's my code snippet:

StockData stockData = new StockData(quotes.Select(x => x.ToTickerData()));
var st = stockData.CalculateSuperTrend(maType: MovingAvgType.ExponentialMovingAverage, length: 10, atrMult: 3.0);

The problem is here:

double prevNvi = nviList.LastOrDefault();
double nvi = currentVolume >= prevVolume ? prevNvi : prevNvi + (prevNvi * pctChg);
nviList.AddRounded(nvi);

because prevNvi is initialized as 0 and therefore nvi also assigned 0 and on next iteration prevNvi again assigned 0 and so on.
solution: prevNvi should never be assigned 0.

Greetings Franklin,

This code

var bars = stockHistoryDaily[@"RELIANCE"];
var stockData = new StockData(bars.Select(x => x.Open), bars.Select(x => x.High), bars.Select(x => x.Low), bars.Select(x => x.Close), bars.Select(x => x.Volume), bars.Select(x => x.DateTime));
var insyncResults = stockData.CalculateInsyncIndex();
var insyncIndex = insyncResults.OutputValues["Iidx"];

produces this exception

OoplesFinance.StockIndicators.Exceptions.CalculationException: Calculations based off of BollingerBands can't be completed because this indicator doesn't have a single output.
   at OoplesFinance.StockIndicators.Helpers.CalculationsHelper.GetInputValuesList(StockData stockData)
   at OoplesFinance.StockIndicators.Calculations.CalculateDetrendedPriceOscillator(StockData stockData, MovingAvgType maType, Int32 length)
   at OoplesFinance.StockIndicators.Calculations.CalculateInsyncIndex(StockData stockData, MovingAvgType maType, Int32 fastLength, Int32 slowLength, Int32 signalLength, Int32 emoLength, Int32 mfiLength, Int32 bbLength, Int32 cciLength, Int32 dpoLength, Int32 rocLength, Int32 rsiLength, Int32 stochLength, Int32 stochKLength, Int32 stochDLength, Int32 smaLength, Double stdDevMult, Double divisor)
   at Program.<Main>$(String[] args) in 

in method CalculateDetrendedPriceOscillator at line CalculationsHelper.GetInputValuesList(stockData).inputList;

      this StockData stockData,
      MovingAvgType maType = MovingAvgType.SimpleMovingAverage,
      int length = 20)
    {
      List<double> doubleList = new List<double>();
      List<Signal> signalList = new List<Signal>();
      List<double> inputList = CalculationsHelper.GetInputValuesList(stockData).inputList;
      int num1 = MathHelper.MinOrMax(checked ((int) Math.Ceiling(unchecked ((double) length / 2.0 + 1.0))));
      List<double> movingAverageList = CalculationsHelper.GetMovingAverageList(stockData, maType, length, inputList);
      int index = 0;

It is called by this code

      List<double> customValuesList5 = stockData1.CalculateBollingerBands(length: length, stdDevMult: stdDevMult1).CustomValuesList;
      List<double> customValuesList6 = stockData.CalculateDetrendedPriceOscillator(length: dpoLength).CustomValuesList;
      List<double> customValuesList7 = stockData.CalculateRateOfChange(rocLength).CustomValuesList;

I am trying out your excellent humongous library. :-)

Hello,

Firstly, thanks for your big Indicator Collection :)

Is there a way you can add TDI ( Traders Dynamic Index ) to your lib?

TDI is a combination of the RSI, BA and MA.

Example From Tradingview > https://www.tradingview.com/script/9Ltwsixb-TDI-Traders-Dynamic-Index-Goldminds/

//
// @author LazyBear
// If you use this code in its orignal/modified form, do drop me a note. 
// 
// Modified by Bromley

study("TDI - Traders Dynamic Index [Goldminds]", shorttitle="TDIGM")

rsiPeriod = input(11, minval = 1, title = "RSI Period")
bandLength = input(31, minval = 1, title = "Band Length")
lengthrsipl = input(1, minval = 0, title = "Fast MA on RSI")
lengthtradesl = input(9, minval = 1, title = "Slow MA on RSI")

src = close                                                             // Source of Calculations (Close of Bar)
r = rsi(src, rsiPeriod)                                                 // RSI of Close
ma = sma(r, bandLength)                                                 // Moving Average of RSI [current]
offs = (1.6185 * stdev(r, bandLength))                                  // Offset
up = ma + offs                                                          // Upper Bands
dn = ma - offs                                                          // Lower Bands
mid = (up + dn) / 2                                                     // Average of Upper and Lower Bands
fastMA = sma(r, lengthrsipl)                                            // Moving Average of RSI 2 bars back
slowMA = sma(r, lengthtradesl)                                          // Moving Average of RSI 7 bars back

hline(30)                                                               // Oversold
hline(50)                                                               // Midline
hline(70)                                                               // Overbought

upl = plot(up, "Upper Band", color = blue)                              // Upper Band
dnl = plot(dn, "Lower Band", color = blue)                              // Lower Band
midl = plot(mid, "Middle of Bands", color = orange, linewidth = 2)      // Middle of Bands

plot(slowMA, "Slow MA", color=green, linewidth=2)                       // Plot Slow MA
plot(fastMA, "Fast MA", color=red, linewidth=2)                         // Plot Fast MA

fill(upl, midl, red, transp=90)                                         // Fill Upper Half Red
fill(midl, dnl, green, transp=90)                                       // Fill Lower Half Green

Hello,

If i use this:

        var rsi = stockData.CalculateRelativeStrengthIndex();

The result are always strange, i only get these numbers: 100 + 50 (60% only 50) and a range from 30 to 50, but nothing between 50.01 to 99.99 and 0 to 29.99, why?

First of all, thank you for this great library :)

I'm considering using SqueezeMomentumIndicator in your library.

In the LazyBear's implementation, there are some input parameters to calculate Bollinger Bands and Keltner Channels,
but your implementation is missing those parameters.

Are those parameters not relevant to the output of the indicator?

I really appreciate any advise you can provide.

I would like to be able to create X number of indicators on the same input data. As per documentation I am using the recomended approach by invoking stockData.Clear();, however, this method seems to be cleaning up the values of the previous outputted collection in some cases (bug?).

Consider following example:

            var results = stockData.CalculateBollingerBands();
            var line2 = results.OutputValues["UpperBand"];
            var line3 = results.OutputValues["MiddleBand"];
            var line4 = results.OutputValues["LowerBand"];

            stockData.Clear();
            var results2 = stockData.CalculateMovingAverageV3();
            var line5 = results2.OutputValues["Mav3"];

            stockData.Clear();
            var results3 = stockData.CalculateSimpleMovingAverage();
            var line6 = results3.OutputValues["Sma"];

            stockData.Clear();

• upon invoke of stockData.Clear(); on the BollingerBands results - the values in collection line2 , line3, line4 are not being cleaned up
• upon invoke of the method 2nd and 3rd times - both the line5 and line6 are affected by the cleaned up as well..

Why the results of the BollingerBands calculation are not being cleaned, while the rest are?

I do think that there is some kind of bug there, since the behavior is expected to be the same for all scenarios.. Proof:

I would like to be able to calculate multiple indicators on the same data set. If the behavior is intended, then how to achieve my goal? (Cloning the resulting list seems to be the temporary solution)