graftonian
1,089 posts
msg #123107
- Ignore graftonian |
3/10/2015 1:54:57 PM
I AM TRYING TO WRITE SOME CODE TO FIND STOCKS WITH A LARGE, FAIRLY EVEN, SINUSOIDAL DIVERGENCE PATTERN. PRESENTLY WORKING WITH CSTDDEV, BUT NO PROGRESS. ANY IDEAS WOULD BE GREATLY APPRECIATED.
THANX,
GRAFTONIAN
|
four
5,087 posts
msg #123110
- Ignore four
modified |
3/10/2015 3:51:22 PM
The Fisher Transformation starts from the analysis of a sine wave.
http://blog.rizm.io/2013/01/the-fisher-transformation-a-trading-system-developed-from-the-pdf-of-the-market-data
|
graftonian
1,089 posts
msg #123111
- Ignore graftonian |
3/10/2015 6:31:46 PM
FOUR,
THANX FOR THE INFO. THIS IS ALL NEW TO ME, SO WILL TRY TO DISSECT IT IN THE MORNING.
GRAFTONIAN
|
Kevin_in_GA
4,599 posts
msg #123112
- Ignore Kevin_in_GA |
3/10/2015 8:17:07 PM
It would be helpful to define exactly what type of divergence you are looking for - I'm assuming it is "the MACD keep going lower while the price rises" sort of thing?
This is easier to filter for as an individual event rather than as a periodic event. For the latter you also would need to define the periodicity of the sinusoidal component as well as what constitutes a "divergence" (how big a difference in percent, for example).
|
graftonian
1,089 posts
msg #123124
- Ignore graftonian |
3/11/2015 1:19:49 PM
FOUR, THANX FOR THE REPLY, LOOKED UP FISHER TRANSFORM ON WIKIPEDIA, FELT LIKE A MONKEY STARING AT A WRIST WATCH (OR MAYBE AN APPLE WATCH).
THINKING AS I WRITE, I ENVISION A SINE WAVE (OR 1/2 SINE WAVE) WITH A PERIOD EQUAL TO THE NUMBER OF DAYS BETWEEN THE TWO POINTS THE MACD VARIANCE IS ZERO. THE AMPLITUDE COULD BE ONE, OR SOME DERIVATION
OF THE VARIANCE AMPLITUDE BETWEEN THE TWO "ZERO" POINTS. THEN DESIGN A FILTER TO MEASURE THE "FIT" OF THE DATA TO THE IDEAL (THE SINE WAVE). I GUESS THE AVERAGE DEVIATION OF EACH DAYS MACD VARIATION TO THE SINE WAVE MIGHT BE A PLACE TO START.
DOES ANY OF THIS MAKE ANY SENSE?
GRAFTONIAN
|
four
5,087 posts
msg #123128
- Ignore four
modified |
3/11/2015 7:25:36 PM
Seems like you might like to investigate John F. Ehlers.
Developed by the author of Rocket Science for Traders
“We measure each stock's dominant cycle period and amplitude to forecast reversals. Note how the Robo-Forecaster accurately predicts turning points in a Theoretical Sine Wave”
— StockSpotter.com Co-Founder John F. Ehlers
http://www.stockspotter.com/In/Books.aspx
|
graftonian
1,089 posts
msg #123131
- Ignore graftonian |
3/12/2015 9:19:41 AM
AS ALWAYS, I'M TRYING TO RE-INVENT THE WHEEL. WILL DO SOME READING ON EHLER TODAY. THANX
|
Mactheriverrat
3,153 posts
msg #123167
- Ignore Mactheriverrat
modified |
3/12/2015 10:49:39 PM
I'll say one thing that IFT(5,9) holding above or below certain levels is very interesting when following up or down trends
Some examples
Submit
|
graftonian
1,089 posts
msg #123182
- Ignore graftonian |
3/13/2015 10:49:13 AM
FOUR,
EHLERS METHODOLOGY IS WHAT I WAS TRYING TO DESCRIBE. HIS BOOKS ARE EXPENSIVE, AND I DON'T KNOW IF I COULD UNDERSTAND THEM AFTER I BOUGHT THEM. I WILL SEE WHAT I CAN LEARN WHILE "PECKING AROUND THE EDGES'.
GRAFTONIAN
|