Topics
Replies
robert.john.edgar
22 Nov 2024, 11:34
If you are really running live with multiple bots that have a 60% win rate and 3+RR then you don't need to ask anyone, you've already got a perfect bot.
The drawdown is more than I would run with but then again your stated returns more than make up for that so I wouldn't worry….
@robert.john.edgar
robert.john.edgar
18 Nov 2024, 20:01
RE: RE: RE: Sharpe Sortino Ratio Missing
PanagiotisCharalampous said:
robert.john.edgar said:
PanagiotisCharalampous said:
Hi there,
Yes it has been removed permanently.
Best regards,
Panagiotis
Seems my local only just upgraded and removed sortino/sharpe.
Why did it get removed?Is there any workaround for this or some plugin or other way we can get this?
My past experience with cTrader optimization is you get pretty poor results if you can't see either sharpe or sortino.
Hi there,
The previous calculation was not correct and the information required to properly calculate these metrics (risk free ratio) is not available in cTrader.
Best regards,
Panagiotis
I would have thought even if acknowledging it was technically wrong they would still maintain the old version as “use at your own discretion” rather than just killing it dead….
It was clear it wasn't producing the expected value but the thing was the value it did calculate was actually really useful at least for a bot on a single asset and a single time period and proved to work in practice.
Sadly without it the optimization maybe not totally useless but is well on the way to being so.
@robert.john.edgar
robert.john.edgar
16 Nov 2024, 19:08
RE: cTrader Ver.5.x has worsened the optimization
dave.anderson.consulting said:
In fact, I think your list of defects is much shorter than reality. I should have looked at the forum earlier. I've been dealing with strange behavior from cBots that have sometimes been in the live account for more than 1 year now for almost a day. Backtesting and Optimization are of course the two regulars that almost never work properly after a release.
I think the UI has deteriorated this time compared to previous releases. I guess they have a new UI designer ;-)Immediately back to previous version 4.9. I won't be using the latest release from now on either. I will stay at least 1 version behind so as not to make my life unnecessarily miserable!
How do you get previous versions?
Same problem as you… but my cTrader is tied to icMarkets
@robert.john.edgar
robert.john.edgar
16 Nov 2024, 17:32
RE: Sharpe Sortino Ratio Missing
PanagiotisCharalampous said:
Hi there,
Yes it has been removed permanently.
Best regards,
Panagiotis
Seems my local only just upgraded and removed sortino/sharpe.
Why did it get removed?
Is there any workaround for this or some plugin or other way we can get this?
My past experience with cTrader optimization is you get pretty poor results if you can't see either sharpe or sortino.
@robert.john.edgar
robert.john.edgar
10 May 2024, 06:17
RE: RE: RE: Push Notifications stopped on iOS
PanagiotisCharalampous said:
robert.john.edgar said:
PanagiotisCharalampous said:
Hi there,
The issue is under investigation and will be resolved asap.
Best regards,
Panagiotis
Thanks,
Notifications started again about 3 hours ago.
Just FYI but assume you are probably aware, the notifications are currently showing UNKNOWN for the asset type i.e. “SELL position for UNKNOWN of 1.00 closed successfully”
CheersRob
Hi Rob,
Does this happen on all symbols or a specific one? Can you share the broker, the symbol and a screenshot?
Best regards,
Panagiotis
Hi there,
Yesterday it was for all asset types.
But when I got up this morning it was now working correctly.
Thanks
Rob
@robert.john.edgar
robert.john.edgar
09 May 2024, 17:24
RE: Push Notifications stopped on iOS
PanagiotisCharalampous said:
Hi there,
The issue is under investigation and will be resolved asap.
Best regards,
Panagiotis
Thanks,
Notifications started again about 3 hours ago.
Just FYI but assume you are probably aware, the notifications are currently showing UNKNOWN for the asset type i.e. “SELL position for UNKNOWN of 1.00 closed successfully”
Cheers
Rob
@robert.john.edgar
robert.john.edgar
02 Jun 2023, 23:43
RE:
pjs.guernsey said:
I have just found this thread whilst searching for the same issue. Same as the OP - ICMarkets allow 0.1 lot sizing on MTx, etc., but in cTrader, it seems to be restricted to 1.0 lots only. Shame, as I signed up specifically to use cTrader.
Yeah I too have come across the same issue.
US30 minimum you can trade through cTrader is 1 lot but icMarkets say you can trade smaller amounts, issue seems specific to cTrader and icMarkets.
Strange they don't bother to fix it.
@robert.john.edgar
robert.john.edgar
23 May 2023, 13:51
RE:
firemyst said:
Good on you for having separate Dev/Production environments :-)
To answer your question, just copy the resulting .algo files over from one machine to the other, and double-click to install.
There's nothing fancy about it. Copy however you want (network share, memory stick, even by attaching to an email and then downloading the attachment on your trading machine)
On your live trading machine, just copy the algo files into the:
1) C:\Users\ [your windows username] \Documents\cAlgo\Sources\Robots for your robots
2) C:\Users\ [your windows username] \Documents\cAlgo\Sources\Indicators for your custom indicators
Thanks for that!
Cheers
Rob
@robert.john.edgar
robert.john.edgar
13 Apr 2023, 13:15
RE:
firemyst said:
Then why not just create your own function that does the same thing?
Should be straight forward.
Just a simple for-loop going through each bar in the data series except the current bar, and compare the values?
Well if you read my original post you will see that's exactly what I am doing, but I had wondered whether I was missing something obvious and was just reinventing the wheel, hence why I posted on this forum to hear from people with more experience at coding cBots about how they were dealing with what should be a common issue.
Rob
@robert.john.edgar
robert.john.edgar
13 Apr 2023, 11:51
RE:
firemyst said:
What do you have for your "Period" setting?
According to the API documentation, the "HasCrossed" should look back the number of bars equal to "period" and see if the values have crossed during that time:
So if you have a period of 1, it should look back over the last 1 bars to see if the values have crossed. If you set period to 2, it should look over the last 2 bars to see if the values have crossed, etc etc.
If not, then please post sample code showing the issue so everyone can see what you're doing and how.
Thank you.
Thanks for the response but I think you slightly misunderstand the API....
A period of 0 looks at only the current bar whilst a period of 1 looks at BOTH the current and the previous bars, and therefore a period of 2 looks at a total of 3 Bars, the current, the previous and the one before that.
Whatever period you use it will ALWAYS include the current unclosed bar in its calculation.
The documentation actually explicitly warns about the problem of these functions when it states:
It is not uncommon that the function will return true and by the end of the bar the two series will uncross.
It about as strong a warning as you can get to not use these functions........
Rob
@robert.john.edgar
robert.john.edgar
07 Apr 2023, 10:29
RE:
clctrader said:
If I'm not mistaken, RMA is the same thing as Wilder Smoothed Moving Average, which is already available.
No it isn't, that's why, for example, in TradingView you can select either WMA or RMA as a MA type.
The main difference between WMA and RMA is in the way they are calculated. WMA assigns weights to each price data point based on the number of periods chosen for the moving average. The most recent price data is given more weight, while older prices are given less weight. This means that the WMA responds more quickly to price changes than the RMA.
On the other hand, the RMA assigns equal weight to each price data point over the specified time period, but adjusts the weight assigned to each data point based on the volatility of the price movement during that time period. This means that the RMA is less sensitive to short-term price fluctuations and is more useful for identifying long-term trends.
@robert.john.edgar
robert.john.edgar
22 Nov 2024, 13:36
RE: RE: RE: RE: Sharpe Sortino Ratio Missing
robert.john.edgar said:
For anyone who might be interested in this topic:
I found this code snippet to calculate the Sortino and Sharpe ratios from an old post.
This will give you the exact same results as Spotware was calculating it.
So cTrader gives us the option to add a custom function into the Optimization Criteria to replace the built in values.
To do this we need to add a GetFitness method to the Bot.
Now we can calculate the old Sortino Ratio we could include it in a custom GetFitness function, for example like this:
From what I had read somewhere I think cTrader just gives each criteria the same weight, but you might want to vary that as above.
Unfortunately there is no way to display the actual Sortino and Sharpe values in the Optimization grid which is unfortunate as personally I used to scan the results.
As ever use at your own risk and always test first….
@robert.john.edgar