| stevehopwoodforex.com https://www.stevehopwoodforex.com/phpBB3/ Print view |
|
| TMA Discussion Thread https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?t=587 |
Page 7 of 17 |
| Author: | garyfritz [ Thu Jun 28, 2012 9:56 pm ] |
| Post subject: | Re: TMA Discussion Thread |
From that post: I wonder if this is the cause of the changing values in the "true" version? Code: Select all Then calcPrevTrue weights bar inx with weight 20, bar inx+1 with weight 21, bar inx+2 with weight 20, etc. That's not an accurate calculation for the previous TMA value. It shouldn't have bar inx with weight 20 in it. Instead of dblTma = calcTmaTrue( symbol, tf, shift ); dblPrev = calcPrevTrue( symbol, tf, shift ); why not do this? dblTma = calcTmaTrue( symbol, tf, shift ); dblPrev = calcTmaTrue( symbol, tf, shift+1 ); That would give dblPrev the accurate previous value of the TmaTrue. |
|
| Author: | NeoTrader [ Fri Jun 29, 2012 5:03 am ] |
| Post subject: | Re: TMA Discussion Thread |
hy gary, hehe...if it where that simple... just because I'm a curious Individual I tested it for you... and this is the result... I also tried/played with different variations and actually the current version is the closest thing that was acceptable from my point of view...and still can be called tma.. It has the same values in the current bar as the TMA and also shows the angle in a clean manner but more compressed and much closer to real time data than the original version. In my understanding of the TMA slope in it's original version the Intention is, that not only the current value of the bar is important to the trader but also the angle of the slope that gives a kind of prediction where the development of the slope is going. Or better described how strong/weak the next bar could become. NB also describes it in his documentation that the stronger the angle is the stronger the move becomes and vice versa. happy trading NeoTrade - |
|
| Author: | Baluda [ Fri Jun 29, 2012 8:15 am ] |
| Post subject: | Re: TMA Discussion Thread |
Because the result would be a LWMA slope instead of a TMA slope. Paul |
|
| Author: | fxlight [ Fri Jun 29, 2012 9:10 am ] |
| Post subject: | Re: TMA Discussion Thread |
Baluda, Neo, can you smooth out TmaSlopeTrue a little with SMA/EMA? Or is it nonsense? fl |
|
| Author: | NeoTrader [ Fri Jun 29, 2012 9:17 am ] |
| Post subject: | Re: TMA Discussion Thread |
hi fxlight It would be nonsense.. We are discussing here against changing the past values to have at least the possibility to do some visual back testing...at least one fraction is against the smoothening... happy trading, NeoTrader - |
|
| Author: | fxlight [ Fri Jun 29, 2012 9:37 am ] |
| Post subject: | Re: TMA Discussion Thread |
Neo, I understand. It is like linear regression. We are looking at current curve (for non fx example). We do not need past values. But fx trader want to see real past values. Not every trader. TmaSlopeTrue is not a very nice curve. So... LOL happy trading |
|
| Author: | NeoTrader [ Fri Jun 29, 2012 10:34 am ] |
| Post subject: | Re: TMA Discussion Thread |
hi fxlight If we want to do quick visual backtest...we need accurate past values and not only FX traders needs them. On the other Hand you can switch anytime to the beautiful original TMA slope...it is only 1 parameter away.. I don't need nice curves... I need accurate curves that resembles the real time values of the past bars. A nice curve brings nothing on the table besides that it looks nice...the information it delivers is faulty. happy trading, NeoTrader - |
|
| Author: | fxlight [ Fri Jun 29, 2012 10:41 am ] |
| Post subject: | Re: TMA Discussion Thread |
thanks, Neo, for your answers. nice weekend fl |
|
| Author: | nonlinear [ Fri Jun 29, 2012 2:28 pm ] |
| Post subject: | Re: TMA Discussion Thread |
Good stuff in this thread. I have been forward testing 10.2, but the inability to review past trades in view of the realtime TMA values is a real pain. Thanks everyone. |
|
| Author: | nonlinear [ Fri Jun 29, 2012 6:58 pm ] |
| Post subject: | Re: TMA Discussion Thread |
I loaded 10.2 TMAslopetrue on all my charts and have been watching the real time values, which look identical to 10.2 TMASlope. The ability to now visually back test is great. |
|
| All times are UTC | Page 7 of 17 |
|
Powered by phpBB® Forum Software © phpBB Limited |
|