SteveHopwood » Mon Jul 31, 2017 5:49 pm wrote:

Johann » Mon Jul 31, 2017 4:14 pm wrote:Hi all,

I have started the test this morning and I might have found a bug with the global basket closure feature.

Nice spot. Thanks.

There is a fix in post 1 - you do not need it if you are not treating every trade on the platform as part of the same basket.

   //Are we in Recovery?
   if (UseRecovery[index] )
   {
      if (MarketBuysCount >= TradesToConstituteRecovery[index])//Minimum trades to constitute Recovery
         if (BuyLoser)//One of them must be a loser or we do not need Recovery
            BuysInRecovery = true;
            
      if (MarketSellsCount >= TradesToConstituteRecovery[index])//Minimum trades to constitute Recovery
         if (SellLoser)//One of them must be a loser or we do not need Recovery
            SellsInRecovery = true;
            
   }//if (UseRecovery[index] )
   

         //Sells
         if (OrderType() == OP_SELL)
         {
            //Recovery
            ArrayResize(SellTickets, MarketSellsCount + 1);
            SellTickets[MarketSellsCount] = OrderTicket();
            //In case the position needs Recovery
           [b] if ( (OrderProfit() + OrderSwap() + OrderCommission()) < 0 )
               SellLoser = true; [/b] 

Johann » Thu Aug 03, 2017 10:04 am wrote:
Thanks for the fix !
I want to report another bloops (hey that is what testing is all about right ??).

The recovery feature for buy trades is not working, but it is only working for the sell trades :

Code: Select all

   //Are we in Recovery?
   if (UseRecovery[index] )
   {
      if (MarketBuysCount >= TradesToConstituteRecovery[index])//Minimum trades to constitute Recovery
         if (BuyLoser)//One of them must be a loser or we do not need Recovery
            BuysInRecovery = true;
            
      if (MarketSellsCount >= TradesToConstituteRecovery[index])//Minimum trades to constitute Recovery
         if (SellLoser)//One of them must be a loser or we do not need Recovery
            SellsInRecovery = true;
            
   }//if (UseRecovery[index] )
   

you can search for the variable "BuyLoser", no value is set outside it init values (set to false).
But the "SellLoser" variable has a value correctly set :

Code: Select all

         //Sells
         if (OrderType() == OP_SELL)
         {
            //Recovery
            ArrayResize(SellTickets, MarketSellsCount + 1);
            SellTickets[MarketSellsCount] = OrderTicket();
            //In case the position needs Recovery
           [b] if ( (OrderProfit() + OrderSwap() + OrderCommission()) < 0 )
               SellLoser = true; [/b] 

This last line of code has been suppressed for the "buys". I have added it again and test running right now !

• a fix for the bug Johan spotted.
• a fix for the non-working negative swap filter. I had coded, resized and initialised all the relevant arrays etc, but had forgotten to code the actual filter.
• a maximum spread filter. You will see the usual MultiplierToDetectStopHunt input underneath the Flying Buddha inputs.
• Baluda sent me the code to internalise the Flying Buddha calculations so Awesome no longer needs the smFlyingBuddha indi. Thanks Paul; that is a fabulous contribution.

SteveHopwood » Thu Aug 03, 2017 10:39 am wrote:I have added a revised version of my Nuclear Option script to post 1. A magic number of -1 tells it to close/delete all trades on the platform, regardless of origin and magic number. Use with care.

Awesome V 1a is in post 1, containing:

• a fix for the bug Johan spotted.
• a fix for the non-working negative swap filter. I had coded, resized and initialised all the relevant arrays etc, but had forgotten to code the actual filter.
• a maximum spread filter. You will see the usual MultiplierToDetectStopHunt input underneath the Flying Buddha inputs.
• Baluda sent me the code to internalise the Flying Buddha calculations so Awesome no longer needs the smFlyingBuddha indi. Thanks Paul; that is a fabulous contribution.

1Direction » Thu Aug 03, 2017 11:24 am wrote:Thanks for the fixes and additional functionality

Can we have a brief discussion about the positive swap filter

As TJF stated yesterday "With the M1 basket set running 0.01 on all pairs I would suggest that keeping it running is fine. If there is a big gap down on one pair there may well be a gap up on another so it should average out in the end."

Now if we use positive swaps=true would this not increase our exposure into one direction of the currencies pairs and increase our global risk exposure (gabs, weekends, news). On the other side the positive swaps has a big impact on the performance specially for higher TF and longer running trades.

My feeling would be to disable positive swaps on the M1 and keep it for the higher time frames as I've only limited experience with +swaps pairs please do correct me if wrong.

Thanks

celte83 » Thu Aug 03, 2017 12:21 pm wrote:Hi
Max DD actually is - 900 pips