Un truc m"échappe sur mon screener de croisement de MM

• 07/01/2017 at 3:36 PM #39547
  laurenzo

  J’essaie de faire un pauvre screener basique 😉

  L’idée c’est juste d’avoir 2 moyennes mobiles différentes en weekly. Quand la plus rapide croise la plus longue, on a notre condition.

  Je colle mon code :

  TIMEFRAME(weekly) //----------------------------// // JMA - Jurik Moving Average // //----------------------------// // --- parameters JMASeries = Close JMAPeriod = 7 // (7 (short), 35 (medium), 175 (long)) JMAPow = 2 // ({ 1..10 }) JMAPhase = 0 // ({-100..100}) // --- beta = 0.45 * (JMAPeriod - 1) / (0.45 * (JMAPeriod - 1) + 2) IF JMAPhase < -100 THEN R = 0.5 ELSIF JMAPhase > 100 THEN R = 2.5 ELSE R = 0.01 * JMAPhase + 1.5 ENDIF IF JMAPow = 1 THEN alpha = beta ELSIF JMAPow = 2 THEN alpha = beta * beta ELSIF JMAPow = 3 THEN alpha = beta * beta * beta ELSIF JMAPow = 4 THEN alpha = beta * beta * beta * beta ELSIF JMAPow= 5 THEN alpha = beta * beta * beta * beta * beta ELSIF JMAPow = 6 THEN alpha = beta * beta * beta * beta * beta * beta ELSIF JMAPow = 7 THEN alpha = beta * beta * beta * beta * beta * beta * beta ELSIF JMAPow = 8 THEN alpha = beta * beta * beta * beta * beta * beta * beta * beta ELSIF JMAPow = 9 THEN alpha = beta * beta * beta * beta * beta * beta * beta * beta * beta ELSIF JMAPow = 10 THEN alpha = beta * beta * beta * beta * beta * beta * beta * beta * beta ENDIF IF BarIndex = 0 THEN Filt0 = JMASeries Filt1 = JMASeries JMA = JMASeries ELSE Filt0 = (1 - alpha) * JMASeries + alpha * Filt0[1] Det0 = (JMASeries - Filt0[0]) * (1 - beta) + beta * Det0[1] Filt1 = Filt0[0] + R * Det0[0] Det1 = (Filt1[0] - JMA[1]) * ((1 - alpha) * (1 - alpha)) + (alpha * alpha) * Det1[1] JMA = JMA[1] + Det1[0] ENDIF //------------------------------------------------------------// // BSF2P - 2-Pole Butterworth Smoothing Filter by John Ehlers // //------------------------------------------------------------// // --- parameters BSF2PSeries = Close BSF2PPeriod = 16 // --- a1 = EXP(-1.414 * 3.14159 / BSF2PPeriod) b1 = 2 * a1 * COS(1.414 * 180 / BSF2PPeriod) coef2 = b1 coef3 = -a1 * a1 coef1 = (1 - b1 + a1 * a1) / 4 IF BarIndex < 2 THEN BSF2P = BSF2PSeries ELSE BSF2P = coef1 * (BSF2PSeries + 2 * BSF2PSeries[1] + BSF2PSeries[2]) + coef2 * BSF2PSeries[1] + coef3 * BSF2PSeries[2] ENDIF //-----------// // Condition // //-----------// Condition = JMA crosses over BSF2P SCREENER[Condition]

  1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86

  TIMEFRAME(weekly)   //----------------------------// // JMA - Jurik Moving Average // //----------------------------//   // --- parameters JMASeries = Close JMAPeriod = 7 // (7 (short), 35 (medium), 175 (long)) JMAPow = 2 // ({ 1..10 }) JMAPhase = 0 // ({-100..100}) // ---   beta = 0.45 * (JMAPeriod - 1) / (0.45 * (JMAPeriod - 1) + 2)   IF JMAPhase < -100 THEN R = 0.5 ELSIF JMAPhase > 100 THEN R = 2.5 ELSE R = 0.01 * JMAPhase + 1.5 ENDIF   IF JMAPow = 1 THEN alpha = beta ELSIF JMAPow = 2 THEN alpha = beta * beta ELSIF JMAPow = 3 THEN alpha = beta * beta * beta ELSIF JMAPow = 4 THEN alpha = beta * beta * beta * beta ELSIF JMAPow= 5 THEN alpha = beta * beta * beta * beta * beta ELSIF JMAPow = 6 THEN alpha = beta * beta * beta * beta * beta * beta ELSIF JMAPow = 7 THEN alpha = beta * beta * beta * beta * beta * beta * beta ELSIF JMAPow = 8 THEN alpha = beta * beta * beta * beta * beta * beta * beta * beta ELSIF JMAPow = 9 THEN alpha = beta * beta * beta * beta * beta * beta * beta * beta * beta ELSIF JMAPow = 10 THEN alpha = beta * beta * beta * beta * beta * beta * beta * beta * beta ENDIF   IF BarIndex = 0 THEN Filt0 = JMASeries Filt1 = JMASeries JMA = JMASeries ELSE Filt0 = (1 - alpha) * JMASeries + alpha * Filt0[1] Det0 = (JMASeries - Filt0[0]) * (1 - beta) + beta * Det0[1] Filt1 = Filt0[0] + R * Det0[0] Det1 = (Filt1[0] - JMA[1]) * ((1 - alpha) * (1 - alpha)) + (alpha * alpha) * Det1[1] JMA = JMA[1] + Det1[0] ENDIF   //------------------------------------------------------------// // BSF2P - 2-Pole Butterworth Smoothing Filter by John Ehlers // //------------------------------------------------------------//   // --- parameters BSF2PSeries = Close BSF2PPeriod = 16 // ---   a1 = EXP(-1.414 * 3.14159 / BSF2PPeriod) b1 = 2 * a1 * COS(1.414 * 180 / BSF2PPeriod)   coef2 = b1 coef3 = -a1 * a1 coef1 = (1 - b1 + a1 * a1) / 4   IF BarIndex < 2 THEN BSF2P = BSF2PSeries ELSE BSF2P = coef1 * (BSF2PSeries + 2 * BSF2PSeries[1] + BSF2PSeries[2]) + coef2 * BSF2PSeries[1] + coef3 * BSF2PSeries[2] ENDIF   //-----------// // Condition // //-----------//   Condition = JMA crosses over BSF2P   SCREENER[Condition]

  Bon ben quand je lance le screener, j’ai pas mal de résultat, mais c’est tout sauf que je veux 😉

  Aucun croisement, rien.

  Un truc doit m’échapper mais je vois pas!

   

  

  laurenzo

  Participant

  Topics: 15

  Replies: 64

  Been thanked: 16 times
  07/01/2017 at 6:33 PM #39560
  laurenzo

  JMA et BSF2P dans le screener n’ont pas les mêmes valeurs qu’avec l’indicateur sur un graphique en weekly.

  Je suis perplexe.

  

  laurenzo

  Participant

  Topics: 15

  Replies: 64

  Been thanked: 16 times
  07/01/2017 at 8:17 PM #39564
  laurenzo

  L’appel à CALL dans le screener fonctionne et me donne une bonne valeur mais quand je copie/colle l’indicateur ça merde. Trop louche sérieux 🙁

  

  laurenzo

  Participant

  Topics: 15

  Replies: 64

  Been thanked: 16 times
  07/02/2017 at 9:57 AM #39571
  GraHal

  I’ve not interpreted / understood your code, but is below okay? Barindex = 0 … means there are no bars??

  IF BarIndex = 0 THEN

  Filt0 = JMASeries

  Filt1 = JMASeries

  JMA = JMASeries

  ELSE

  1 user thanked author for this post.

  laurenzo

  

  GraHal

  Participant

  Topics: 211

  Replies: 8764

  Been thanked: 2179 times
  07/02/2017 at 12:38 PM #39583
  laurenzo

  It’s okay, the index of BarIndex starts to 0. That’s the first bar to the left of the chart.

  In fact I have just copied/pasted the code of my indicators in the screener and I have not the same results (wrong ones) than when they are attached on a chart in weekly.

  It only works when I use CALL.

  It’s very strange.

   

  

  laurenzo

  Participant

  Topics: 15

  Replies: 64

  Been thanked: 16 times
  07/04/2017 at 9:38 AM #39760
  GraHal

  Post your code with the indicator pasted in (no CALL) and I’m sure somebody will spot the issue?

  Nous devrions revenir au français ici encore, Nicolas nous “aura-t-il”! 🙂

  

  GraHal

  Participant

  Topics: 211

  Replies: 8764

  Been thanked: 2179 times
  07/04/2017 at 9:50 AM #39761
  Nicolas

  Merci GraHal pour le rappel à l’ordre ! 😀

  En effet lorsque tu CALL l’indicateur, celui-ci ne doit pas être interprété par le moteur de ProScreener mais par celui de ProBuilder et les résultats peuvent différer. Même si ça ne semble pas être le cas ici, il ne faut pas oublier que ProScreener a une limite de 254 bars d’historique, peu importe le timeframe utilisé.

  

  Nicolas

  Keymaster

  Topics: 81

  Replies: 20863

  Been thanked: 4773 times
• Author
  Posts