You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
DMI/ADX consistently works
- Thread starter nyc-hotshot
- Start date
WarEagle,
Here is another formula for DX:
http://www.aspenres.com/Documents/AspenGraphics4.0/Average_Directional_Index.htm
It is different than what you are calculating.
This says DX = 100 *(+DI-(-DI) / +DI+(-DI))
It does not use absolute values....maybe try that out.
Here is another formula for DX:
http://www.aspenres.com/Documents/AspenGraphics4.0/Average_Directional_Index.htm
It is different than what you are calculating.
This says DX = 100 *(+DI-(-DI) / +DI+(-DI))
It does not use absolute values....maybe try that out.
Quote from Stok:
WarEagle,
Here is another formula for DX:
http://www.aspenres.com/Documents/AspenGraphics4.0/Average_Directional_Index.htm
It is different than what you are calculating.
This says DX = 100 *(+DI-(-DI) / +DI+(-DI))
It does not use absolute values....maybe try that out.
The multiplication by one hundred doesn't matter since this scale is of percentages of 1 - it would just force it to a scale of 0-100 instead of 0-1.
It still needs the absolute value in that calcuation since we aren't noting any direction, just strength.
That said, I have been trying a variety of different moving averages and incorrect ADX calculations, including the one above, in an Excel spreadsheet and I'm still not getting what it gets on screen.
It is very close to just having straight up DX.
Either way if it works on that system, it is good to know - I will try to find some time to test that system to see if I can get it to work for me.
Quote from Rock1:
Hello, nyc-hotshot
I did not at all intend to criticize your system or doubt about the signals. It is really great that you are willing to share your experiences. I simply wanted to give this information because when somebody uses a different charting software which calculates the real ADX after the guidelines of its originator Welles Wilder, then she/he will not be able to reproduce your signals.
best regards
Rock1
No problem Rock, I was not referring to you anyway. Also, I don't mind debate on the subject. That's what the boards are for. From what you said I assume you are right and the indicator is probably DX not ADX. I can accept that.
It is the people who made stupid allegations about me "spamming" and others who said they "seriously question" things I'm saying. That is just annoying and uncalled for. I am no technical or math wizard, and I have never professed to be. I am a trader, who has day traded since 1996, and has made good money.
I just wanted to share what works for me (charts, trades, etc...) and let people make their own determinations about whether it will work for them. I think if you don't spell out every little detail, then some poeple think you are full of it, but the charts don't lie.
Stok,
Yeah, in that formula you gave, the *100 won't really matter except for where the decimal is. The absolute value is used to keep the values above zero. If you plot the one without the absolute value, you will get a line that fluctuates around zero. nyc's DX line does not go negative, so it is using absolute value in the calculation. Thanks for the heads up though.
Yeah, in that formula you gave, the *100 won't really matter except for where the decimal is. The absolute value is used to keep the values above zero. If you plot the one without the absolute value, you will get a line that fluctuates around zero. nyc's DX line does not go negative, so it is using absolute value in the calculation. Thanks for the heads up though.
Quote from AsaFce:
It still needs the absolute value in that calcuation since we aren't noting any direction, just strength.
I don't use absolute value when I caculate ADX. I just do:
DX = (pdi - mdi) / (pdi + mdi)
ADX = emov(DX)
Seems to be correct to me, FWIW. Can't remember where I original came across the formula, but that's what it was.
Quote from phoenix_rising:
I don't use absolute value when I caculate ADX. I just do:
DX = (pdi - mdi) / (pdi + mdi)
ADX = emov(DX)
Seems to be correct to me, FWIW. Can't remember where I original came across the formula, but that's what it was.
So you are saying that you have no qualms with there being a negative DX value?
I've seen an example of it on I think optionstrader.com or something that didn't have the absolute value reference in there, but it seems an oversight to me.
It is a directional indicator, and the name (pdi - plus, mdi - minus) denotes the direction. The value is solely the strength.
Meaning that you shouldn't see a negative number.
Much in the same way that you can't drive negative 55mph.
OK, I take that back. I just went back and looked at my source. I do take the ABS of the numerator. The denominator is always positive, so that one doesn't need it. My bad.
Yeah...if you don't use AV you will get negative values.
Example
DM+ = .10
DM- = .20
Formula:
DX = (.10-.20) / (.10+.20) will give you -.10/.30 - thats a neg #.
LOL, this whole thing is kinda funny, because nyc's methods look awesome to me and it might come from a funky ADX line this is programmed wrong/differently or something.....gonna try to keep working on it as well guys.
Stok
Example
DM+ = .10
DM- = .20
Formula:
DX = (.10-.20) / (.10+.20) will give you -.10/.30 - thats a neg #.
LOL, this whole thing is kinda funny, because nyc's methods look awesome to me and it might come from a funky ADX line this is programmed wrong/differently or something.....gonna try to keep working on it as well guys.
Stok