Variable and figures concatenation

Roger · 2020-01-28T18:37:46+00:00

Hi, I would like to try optimisation tests based on the time. I can define a trading hour to enter long or short but I would like to insert 2 variables in my trading times in order to optimize : defparam cumulateorders = false TradingHour = Time = a'0000' TradingHourOut = Time = b'0000' condenterlong=close<close[1] and TradingHour condexitlong=TradingHourOut IF NOT LongOnMarket AND condenterlong THEN BUY 1 CONTRACTS AT MARKET ENDIF If LongOnMarket AND condexitlong THEN SELL AT MARKET ENDIF As you can see my variables are a and b Thanks

Viewing 6 posts - 1 through 6 (of 6 total)

Author

Posts
01/28/2020 at 7:37 PM #118175 quote
Roger
Participant

Veteran
Hi,

I would like to try optimisation tests based on the time.

I can define a trading hour to enter long or short but I would like to insert 2 variables in my trading times in order to optimize :
```
defparam cumulateorders = false

TradingHour = Time = a'0000'
TradingHourOut = Time = b'0000'

condenterlong=close<close[1] and TradingHour
condexitlong=TradingHourOut


IF NOT LongOnMarket AND condenterlong THEN
BUY 1 CONTRACTS AT MARKET
ENDIF


If LongOnMarket AND condexitlong THEN
SELL AT MARKET
ENDIF
```
As you can see my variables are a and b

Thanks
01/28/2020 at 8:41 PM #118178 quote

nonetheless
Participant

Master

What I would do is:

TradingHour = a

TradingHourOut = b

and then set the range in the optimisation box, for example:

a 070000 to 100000 in steps of 10000

b 150000 to 210000 in steps of 10000

Unless I’ve misunderstood what you want to achieve?

01/28/2020 at 8:54 PM #118180 quote

Roger
Participant

Veteran

That is exactly what I want to achieve.

I have no computer tonight but I will try it tomorrow.

Thanks for your help!

01/29/2020 at 7:22 PM #118221 quote

Roger
Participant

Veteran

Your solution seems to work properly, thanks again, I am going to do some “curve fitting” 🙂

01/29/2020 at 8:56 PM #118223 quote

Vonasi
Moderator

Master

Roger wrote:

I am going to do some “curve fitting” 🙂

If only I’d known to buy on a Tuesday at 1432 and sell on Wednesday at 1559 then I’d be a rich man.

At least you are going into this study knowing that there is a huge potential for data mining when using time as a variable. 🙂

01/29/2020 at 10:06 PM #118226 quote

Roger
Participant

Veteran

I have spent more than one hour on it and even with overfitting I can barely reach a profit factor of 2 and more than 50% of winning trades… The time window 17h00-18h00 seems to be the most interesting window to work on but for now it is extremely disappointing…
Author

Posts