Posts made by backtrader

You post code which says you are doing this:

@emmepra said in Plot BuySell Observer with Tick Data:

cerebro = bt.Cerebro(stdstats=False)
cerebro.addobserver(bt.observers.BuySell)

But the chart shows you are NOT doing that, because the observer has the following legend: BuySellObserver(True, 0), where True is the value for the parameter barplotdist, which has a default value of False. Furthermore, the parameter bardist has a default value of 0.015 (i.e.: 1%) but the chart shows it to be 0.

Two parameters have changed its default value as shown in the chart, but the code which supposedly doesn't work doesn't change anything.

You may actually want to try with the code you show (I never tried it with ticks ... it may or it may not work)

Use the len of the data to know how many items are available.

@t3chap said in How to build a sizer that can use some measure of volatility, like ATR, and the multiplier for the future to determine position sizing?:

I seem to have a position size of 0 for every trade.

With the following code ...

@t3chap said in How to build a sizer that can use some measure of volatility, like ATR, and the multiplier for the future to determine position sizing?:

class comm_sizer(bt.Sizer):
    def __init__(self, atr):
        #self.atr_parameter= atr_parameter
        atr= self.atr
        risk_per_trade=5000
        atr_risk_per_trade=5
        
        sizing = math.floor((risk_per_trade/(mult*atr_risk_per_trade*atr)))
        
   # position = self.broker.getposition(data)
    
    def _getsizing(self,comminfo,cash,data,isbuy):
        
        return sizing

It would be a miracle if you get something other than an Exception, both in __init__ and in _getsizing.

In any case to use the ATR you probably want to access the strategy attributes (where the ATR will have been instantiated), taking advantage of the fact that the sizers have an attribute member called strategy

• https://www.backtrader.com/docu/sizers/sizers/

Dear @P-dML ,

For sure your explanation makes sense to you, because you are deep into your problem. But as pointed out by @run-out in the question ... without code there is actually no way to know what you are doing. For example:

• There is a getsizing(data) thrown into a sentence out of the blue ...

At least for me it is unclear why that isn't giving you the information you are missing, probably because I also don't know which data feeds are in the equation, what's being targeted, ...

@Bayne said in Memory Error Plotting (Ticks):

Data is not more than 2 weeks, but even need more.

Buy a PC with more memory and you will be able to plot more (OS may impose limits on memory allocation)

@Bayne said in Memory Error Plotting (Ticks):

Is it possible to plot it in Upsampled Bars, to save memory?

It will obviously use less room. But you won't see the ticks.

@Bayne said in Does it pay off (in accuracy terms) to load lower TimeFrame Data?:

Do upsampled bars Take into consideration the smaller moves in between (which are Not OHLC) for executing Orders or not?

You obviously don't now what upsampling (known as resampling) means. Smaller moves cannot be taken into account because the information is lost. You compress (with information loss) n-bars into 1-bar.

You may use use the resampled data to trigger signals and target the lower timeframe for execution. Or use replay: https://www.backtrader.com/docu/data-replay/data-replay/

You can always use the non-adjusted close and see how your strategy fares when a 5x1 split happens.

You are for sure NOT doing that.

@emmepra said in Plot BuySell Observer with Tick Data:

cerebro = bt.Cerebro(stdstats=False)
cerebro.addobserver(bt.observers.BuySell)

If you were the observer wouldn't say ... BuySell(True, 0). That's the problem with redacted non-working code, it is usually not the code actually executing.

See: https://www.backtrader.com/docu/observers-reference/

@jesseliu0 said in How to disable plotting data?:

only plot observers and analyzers?

analyzers are not plotted in any instance.

It is unclear from your message what happens when you run the code.

What do you have to contribute?

@FaiqS said in Multiple Datafeeds and multiple Timeframes:

Which is not convenient to iterate through in next function of strategy.

It may not be convenient for you. Don't mix them in such a way then. Add first the lower timeframes and later the higher timeframes. You will have two clear partitions.

Or iterate using steps other than 1 and with different starting indices.

@FaiqS said in Multiple Datafeeds and multiple Timeframes:

The straightforward it is referring to is to add "cerebro.adddata(data)"

No. It is referring to resampledata(data, ...). Using any other approach won't change the order in which things are in the system and where.

1. It's not an issue and the article doesn't say so. It says that people tend to believe it is an issue but it is lack of understanding.

2. You are NOT getting duplicate records. But you fail to show what you print, so it is virtually impossible to say what you are printing.

3. When a data feed moves past other and the "other" has no new records only the existing price can be shown.

4. Data feeds (and mostly everything) has a len to indicate if something new has been added. Check the len.

Lines do only store floats. Nothing prevents that you store more complicated structures as member attributes of your indicator.

How much do you pay for the project?

Let me guess: NOTHING

How much are you willing to pay for continued updates that satisfy your needs?

Let me guess: NOTHING.

What is what you have to contribute?

Let me guess: NOTHING

What have you found to be missing?

Let me guess: NOTHING

Please read this: https://www.backtrader.com/blog/posts/2016-04-19-sync-different-markets/sync-different-markets/

Some people tend to name the script backtrader.py and end up importing the script and not the package.

@edoardo-mancinelli said in Time axis missing in the plot:

any advice?

Yes. Search ... https://community.backtrader.com/topic/2016/basic-plotting-no-x-axis

FIRST LESSON IN ALGOTRADING (believe me when I say it is an important one)

In algotrading like in chess, look carefully at what's right before your eyes, like what's at the top of every page of this community.

For code/output blocks: Use ``` (aka backtick or grave accent) in a single line before and after the block. See: http://commonmark.org/help/

2nd, this thread says (one example)

@backtrader said in Plotting only chosen data?:

@run-out said in Plotting only chosen data?:

You do this by setting plot=False when creating your dataline.

This can be done at anytime, not only during creation.

The same can be read in the documentation here: https://www.backtrader.com/docu/plotting/plotting/

Where are you doing that in your code which seems fairly incomplete? (Actually for any of the data, you have many and only 2 are shown)

You are not even storing the resampled data instance in a variable to do that

Nobody is interested in your secret sauce. If you want to show something and hide your very profitable code, create a 10 line snippet to show the problem.

Let us see what you do ... to find out why you cannot.

The default configuration can be switched off when running cerebro.run with the parameter stdstats.

• https://www.backtrader.com/docu/cerebro/

After that you can add your own plethora of observers, including the BuySell one and decide where you want the markers

• https://www.backtrader.com/docu/observers-and-statistics/observers-and-statistics/
• https://www.backtrader.com/docu/observers-reference/

An example

• https://www.backtrader.com/blog/posts/2016-12-10-buysellarrows/buysellarrows/