Support Board
Date/Time: Thu, 16 Jan 2025 13:15:28 +0000
[User Discussion] - Ehlers Roofing filter
View Count: 2642
| [2014-12-22 22:13:31] |
| ehlaban - Posts: 50 |
|
From TASC Magazine Jan 2014. Ganz has also uploaded a version. Thanks Ganz i learned a lot from your code! [usr dscssn] Roofing filter study |
 ehlEhlersRoofingFilter.cpp - Attached On 2014-12-22 22:12:30 UTC - Size: 2.79 KB - 533 views ehlEhlersRoofingFilter.dll - Attached On 2014-12-22 22:12:43 UTC - Size: 16.5 KB - 505 views |
| [2017-06-21 13:41:56] |
| michelh - Posts: 159 |
|
thanks ehlaban, just a quick question - I know that your code is correct but i have a question about coding concepts. his definition of alpha (TS code) is: alpha1 = (Cosine(.707*360 / HPPeriod) + Sine (.707*360 / HPPeriod) - 1) / Cosine(.707*360 / HPPeriod); Your code reads: float tr, fi; tr = 360 * 3.14159f / 180; fi = 0.707f * tr / Bars.GetInt(); alpha1 = (cos(fi) + sin(fi) - 1) / cos(fi); so you have introduced an extra factor of 2 pi into the sin/cos in defining alpha. Why did you have to do this? And why didn't you write tr = 2 * 3.14f? why did you do the 360/180? When I coded it myself, I got nothing but gibberish in the output. But if you have 2 minutes to write me a note to tell me why you had to do it like this, that would be appreciated. Many thanks, Michel |
| [2017-06-21 18:03:14] |
| ehlaban - Posts: 50 |
|
Mmm, that's a long time ago. I think it has to do with the conversion from radians to degrees https://stackoverflow.com/questions/135909/what-is-the-method-for-converting-radians-to-degrees https://gist.github.com/dbrockman/4773781 http://www.cplusplus.com/forum/beginner/197655/ |
| [2017-11-09 18:01:55] |
| Jezz - Posts: 33 |
|
One Time Calculations That Do Not Run During Study Updating Just watching the code, I realized that filter coefficients are calculated every updating time. Filter coefficients need to be calculated just once from the period parameter. example from the supersmoother section a1 = exp((-1.414f) * 3.14159f / LPPeriod.GetInt());
b1 = 2 * a1 * cos(1.414f * (180 * 3.14159f / 180) / LPPeriod.GetInt()); c2 = b1; c3 = (-1)*a1*a1; c1 = 1 - c2 - c3; All these values depends only on LPPeriod, so they need to be calculated once after the "period" parameter change is applied. I've tried to move this portion of code to a if section as suggested in ACSIL Programming Concepts: One Time Calculations That Do Not Run During Study Updating without any success. If I try to change the value of period, it doesn't look to affect. if(sc.Index == 0)
{ //Code to run only on study full recalculation .... } Is this the right approach? Any suggestion welcome Thanks |
To post a message in this thread, you need to log in with your Sierra Chart account: