Renko automatic trading

• 12/11/2020 at 12:23 PM #153364
  eckaw

  Since we have really limited backtest data on 10s timeframe: would it be an idea to optimise every 2 days, pick 10 results then take an average over a week or month? Or optimise daily and take an average after a couple of weeks?

  Any thoughts on how to get an average optimised value on this timeframe that might actually work?

  

  eckaw

  Participant

  Topics: 13

  Replies: 118

  Been thanked: 61 times
  12/11/2020 at 12:25 PM #153366
  eckaw

  @snuckle would you mind sharing your version here?

  

  eckaw

  Participant

  Topics: 13

  Replies: 118

  Been thanked: 61 times
  12/11/2020 at 2:31 PM #153404
  snucke

  i just replaced your indicator with this one.

  your idea on getting an average over a couple of weeks is not a bad idea, maybe reoptimize after 1 week and keep the first still live and see how the new one performs

  and at the same time take an average of the first optimization and the second and at the end have 3 running at the same time

  

  snucke

  Participant

  Topics: 9

  Replies: 143

  Been thanked: 21 times
  12/11/2020 at 4:35 PM #153425
  Paul

  thnx. I’ve got a snippet I added to a rough version, i’am unsure how useful it will be.

  What it does i.e. for long, it shows the buy signals you miss when you’re in a long positionposition, So it doesn’t count the signal when the position is opened, only the buysignals when longonmarket.

  It does also calculated cumulative those signals which total could go up quite rapidly.

  In general the higher the amount of cumulative signals, is not a good sign for robustness.

  for testing if interested.

  conbuy=buysignal condsell=sellsignal // show signals once showsignals=1 if showsignals then once count1l=0 once count1s=0 if not onmarket or ((longonmarket and shortonmarket[1]) or (longonmarket[1] and shortonmarket)) then count2l=0 count2s=0 endif if longonmarket then if condbuy then count1l=count1l+1 count2l=count2l+1 endif endif if shortonmarket then if condsell then count1s=count1s-1 count2s=count2s-1 endif endif graph count1l as "long signals in position cumulative" graph count2l as "long signals in position" graph count1s coloured(255,0,0,255) as "short signals in position cumulative" graph count2s coloured(255,0,0,255) as "short signals in position" endif

  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

  conbuy=buysignal condsell=sellsignal   // show signals once showsignals=1 if showsignals then once count1l=0 once count1s=0   if not onmarket or ((longonmarket and shortonmarket[1]) or (longonmarket[1] and shortonmarket)) then count2l=0 count2s=0 endif   if longonmarket then if condbuy then count1l=count1l+1 count2l=count2l+1 endif endif if shortonmarket then if condsell then count1s=count1s-1 count2s=count2s-1 endif endif   graph count1l as "long signals in position cumulative" graph count2l as "long signals in position"   graph count1s coloured(255,0,0,255) as "short signals in position cumulative" graph count2s coloured(255,0,0,255) as "short signals in position" endif

   

   

   

  4 users thanked author for this post.

  icharttop, eckaw, GraHal

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  12/11/2020 at 7:26 PM #153430
  GraHal

  Link to above added as Log 263 to here …

  Snippet Link Library

  3 users thanked author for this post.

  eckaw, Paul

  

  GraHal

  Participant

  Topics: 204

  Replies: 8423

  Been thanked: 2045 times
  12/11/2020 at 9:45 PM #153438
  eckaw

  @paul by implementing your snippet, do we need to change cumulateorders= true? I was experimenting with cumulate orders = true in my code as you can see but failed to get more positive results. I think it would be worth taking a look at implementing some reversal code to exit trades as I think the current one isn’t performing very well.

  

  eckaw

  Participant

  Topics: 13

  Replies: 118

  Been thanked: 61 times
  12/13/2020 at 9:33 AM #153540
  Paul

  eckaw wrote:

  do we need to change cumulateorders= true

  no because although it hits the same points, I wanted to optimise a strategy normally with 1 position, but still have visible the missed signals & the totals.

  I’am not a fan of reversal because it cover ups the fact that the original signal was poor.

  1 user thanked author for this post.

  eckaw

  

  Paul

  Participant

  Topics: 40

  Replies: 812

  Been thanked: 660 times
  12/14/2020 at 8:08 PM #153751
  supermyguel

  snucke wrote:

  i just replaced your indicator with this one.

  Hi!

  If not too complicated can you write how replace the indicator? I have to modify some parameters? If I simply replace there is an error about number of called parameters….

  Thanks in advance!

   

  

  supermyguel

  Participant

  Topics: 13

  Replies: 60

  Been thanked: 1 time
  12/18/2020 at 1:30 PM #154279
  snucke

  Results from first optimization on 70/30 WF have now been live for one week, seems ok so far.

  did a new optimization on 70/30 WF and will be running it live simultaniously to this one and see if the first one fades or not.

  1 user thanked author for this post.

  GraHal

  

  snucke

  Participant

  Topics: 9

  Replies: 143

  Been thanked: 21 times
  12/18/2020 at 3:31 PM #154302
  GraHal

  supermyguel wrote:

  If not too complicated

  Maybe snucke would prefer to share his version of the .itf with us all?

  Be easier than typing instructions etc??

  

  GraHal

  Participant

  Topics: 204

  Replies: 8423

  Been thanked: 2045 times
  12/18/2020 at 3:53 PM #154309
  snucke

  sure

  

  snucke

  Participant

  Topics: 9

  Replies: 143

  Been thanked: 21 times
  12/18/2020 at 6:00 PM #154327
  DjungelJarl

  Hi snucke and everyone else, how do you ensure that the minimum stop distance isn’t violated? I’m on IG and I’ve been trying a few different versions but the algorithm fails because the minimum stop distance of 6 pips isn’t respected. Thanks in advance for any tips regarding this.

  

  DjungelJarl

  Participant

  Topics: 0

  Replies: 13

  Been thanked: 2 times
  12/18/2020 at 6:25 PM #154331
  nonetheless

  IG minimum stop distance on the DJ is 10

  

  nonetheless

  Participant

  Topics: 83

  Replies: 1237

  Been thanked: 978 times
  12/18/2020 at 6:43 PM #154334
  DjungelJarl

  Right, my bad. Any idea of how to configure this in the code?

   

  EDIT: This restriction applies to the stop orders as well right?

  

  DjungelJarl

  Participant

  Topics: 0

  Replies: 13

  Been thanked: 2 times
  12/18/2020 at 7:00 PM #154336
  DjungelJarl

  This is my attempt at respecting the stop distance, it was added in the end of the stop loss code in snuckes latest version of this. I’m putting it on demo again to see if it works.

  Attempt at respecting minimum stop distance

  if newSL > 0 then //and abs(newSL-tradeprice) < stopRoom then if newSL-tradeprice > 0 and newSL-tradeprice < stopRoom then newSL = newSL + (stopRoom-(newSL-tradeprice)) elsif newSL-tradeprice < 0 and newSL-tradeprice > -stopRoom then newSL = newSL - (stopRoom+(newSL-tradeprice)) endif endif

  1 2 3 4 5 6 7

  if newSL > 0 then //and abs(newSL-tradeprice) < stopRoom then if newSL-tradeprice > 0 and newSL-tradeprice < stopRoom then newSL = newSL + (stopRoom-(newSL-tradeprice)) elsif newSL-tradeprice < 0 and newSL-tradeprice > -stopRoom then newSL = newSL - (stopRoom+(newSL-tradeprice)) endif endif

   

  

  DjungelJarl

  Participant

  Topics: 0

  Replies: 13

  Been thanked: 2 times
• Author
  Posts