Error: Unknown Command. Code is invalid. Please correct it line 8:

• 06/13/2023 at 12:43 AM #216040
  Pieter_T

  Good point Peter. Yes, that was my first thought as well. In the indicator, the parameter is definitely set up as “Decimal”. If I use the indicator on the chart directly and use decimal values it works correctly. The backtesting also works fine using decimal values. It is just when you select the prepare for auto-trading that for some reason it does not want to accept the decimal values. Did you have this issue before? The workaround I use now is to use whole numbers in the parameter setting and then divide the whole numbers in the code to get the decimal places I need. That works fine but it is still an annoyance.

  

  Pieter_T

  Participant

  Topics: 5

  Replies: 27

  Been thanked: 5 times
  06/13/2023 at 3:33 AM #216045
  PeterSt

  pietertal wrote:

  The workaround I use now is to use whole numbers in the parameter setting and then divide the whole numbers in the code to get the decimal places I need.

  Haha, yes, I saw you doing that. This is why/how I came to the question.
  But no, I don’t recognize that and using decimals via the normal Optimization Parameters I am doing all the time (and they work out).
  I also recognise the /10000 which I use myself regularly, but this is because the numbers in the Optimization Parameters can’t get small enough. From the top of my head : 4 is the maximum while more normal would be 5 (which is the max for Rounding) and beyond of course. What I mean with this is :

  When you iterate from 0.00001 to 0.00010, this all shows the same in the Result List. Still internally it *will* use that 5th decimal but as you can’t see it, it is useless.
  Can’t this be your problem ?
  It is even hard to recognise for me because when such a thing happens, I start to obtain the iteration as 1-10 and divide by 10000 … I do this regardless. And with Forex this is a nobrainer. In the 1st attachment you can see that the decimals are used all over the place, but these are the Optimization Parameters only (I don’t use CALLable Signals). In the 2nd attachment you see how I wouldn’t run into the problem because of the standard approach for Fx (vs not-Fx).

  If you think it is worth the trouble, you could make a Technical Report of it (goes via the Help menu).
  And redundancy is never a good thing.

  1 user thanked author for this post.

  Pieter_T

  

  PeterSt

  Participant

  Topics: 71

  Replies: 2313

  Been thanked: 699 times
  06/13/2023 at 3:56 AM #216049
  Pieter_T

  Just came to realise something looking at your first attachment. The smallest decimal you use in your example is 0.0011. I used this value as a test and it has no error. If I however change it to 0.0001 the error is back. I can however use 0.0011111 and there is no error. In summary, what I realise is that there cannot be more than 3 zero’s after the decimal point.

  I will submit a report and see if something comes of it.

  1 user thanked author for this post.

  PeterSt

  

  Pieter_T

  Participant

  Topics: 5

  Replies: 27

  Been thanked: 5 times
  06/13/2023 at 2:21 PM #216123
  GraHal

  pietertal wrote:

  there cannot be more than 3 zero’s after the decimal point.

  I agree as I’ve had the same experience with < 0.0001

  1 user thanked author for this post.

  Pieter_T

  

  GraHal

  Participant

  Topics: 204

  Replies: 8420

  Been thanked: 2044 times
• Author
  Posts