Stoch & KDJ

[muteddisk]

Hi,
Good work on your code. Would you mind adding KDJ to your indicators? KDJ is Stoch with added J. The J line represents the divergence of the D value from the K value

thanks

[DaveSkender]

If %J really is just a weighted difference between Stochastic Oscillator %K and %D, I'll likely just add it to the output of Stoch and make reference to KDJ in the documentation. This is a similar technique used in MACD and PVO. In the meantime, here's a suggestion on how to get %J in your wrapper code.

// your custom derived class
public class KdjResult : StochResult
{
  public decimal PercentJ { get; set; }      // %J
}

[ .. ]

List<KdjResult> results = Indicator.GetStoch(history,9,3,3);
  .Select(x => new KdjResult
    {
      Date = x.Date,
      Oscillator = x.Oscillator,              // %K
      Signal = x.Signal,                      // %D
      PercentJ = 3*x.Oscillator - 2*Signal    // %J
    })
  .ToList();

[DaveSkender]

I'm looking for good references for the formula and I see most are using the standard Slow Stochastic values 9,3,3 as in my example above, then use %J=3*%K-2*%D or %J=3*%D-2*%K -- not sure which is correct yet. I'm also wondering if those 3 and 2 weights should be derived from those smoothing factors or if they are truly static values? I suspect they are static. 🤔

If anyone has any legitimate documentation on this from the original author, George Lane, please share.

[DaveSkender]

I'm seeing both of these %J calculations used in various charting platforms on the web. I've not found the original author's formula, because I can't find it (please share if you do); however, I'm going with %J=3*%K-2*%D because it's consistently depicted this way in a couple of academic research articles that I've found. I also suspect this is correct since it depicts more of a leading indicator.

For comparison, here's %J=3*%D-2*%K (incorrect?):

And here's %J=3*%K-2*%D (this is the one I'm using):

[DaveSkender]

Does anyone know how to trade this indicator? Please explain if you're experienced with it -- do not just repost information and links from random blogs. I'm curious to know how this is different from the simple %K / %D crossover signaling model.

[DaveSkender]

This is now available in version 1.6.3 for the GetStoch method. See Stochastic Oscillator for more information.

[muteddisk]

Thanks bud your the best.

…

On Sun., Jan. 10, 2021, 13:14 Dave Skender, ***@***.***> wrote: If %J really is just a weighted difference between Stochastic Oscillator %K and %D, I'll likely just add it to the output of Stoch and make reference to KDJ in the documentation. This is a similar technique used in MACD and PVO. In the meantime, here's a suggestion on how to get %J in your wrapper code. // your custom derived classpublic class KdjResult : StochResult { public decimal Histogram { get; set; } // %J } [ .. ] List<KdjResult> results = Indicator.GetStoch(history); .Select(x => new KdjResult { Date = x.Date, Oscillator = x.Oscillator, // %K Signal = x.Signal, // %D Histogram = 3*x.Oscillator - 2*Signal // %J }) .ToList(); — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#315 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AHFLO6YZDIVELL6MR4KKCQDSZHVA7ANCNFSM4V4P6QLQ> .

[wclark17]

Just curious if you could add an alias of K & D to Oscillator and Signal? Totally not important, but helps my brain at times.

BTW, You are saving me a ton of work elsewhere. Happy to sponsor you and will increase it soon!

[DaveSkender]

I’ll take a look, track on #569. I’ve had this same conversation in my head a few times, both ways, “what is D again?” I settled on the more descriptive variants, but can see how others like the other naming.

let me see if I can add aliases in a way that doesn’t cause more potential confusion.

[DaveSkender]

and thank you for your patronage, much appreciated