Technically it should work ( in start method for example), however IMHO it is better to set the position outsize the strategy code - just in case you would like to re-use your strategy in live-trading or otherwise keep you strategy code independent of external setup.
Analyzer is probably a better place - you may always switch the analyzer on for back testing and off for live trading for example - leaving your strategy code intact.
A wild guess may be that most of the time is wasted loading the data. I would measure it with timeit:
start_time = timeit.default_timer()
for i in range(len(self.data_list)):
data = PandasData(
dataname=self.data_list[i][0], # Pandas DataFrame
name=self.data_list[i][1] # The symbol
)
cerebro.adddata(data)
elapsed = timeit.default_timer() - start_time
print(f'loading data: {elapsed}')
#cerebro.addanalyzer(ScreenerAnalyzer)
start_time = timeit.default_timer()
cerebro.run()
elapsed = timeit.default_timer() - start_time
print(f'run: {elapsed}')
It seems the link in the original topic is no longer valid. Here is the updated link: https://www.backtrader.com/blog/2019-05-20-momentum-strategy/momentum-strategy/
IMHO the better way is to define the CP2 line using delaying indexing. See Delayed Indexing Concept.
Probably something like:
def __init__(self):
self.addminperiod(self.params.periodSlow)
# Calculate moving averages
self.movavSlow = self.p.movav(self.data.close, period=self.p.periodSlow)
self.movavFast = self.p.movav(self.data.close, period=self.p.periodFast)
# Perry Kaufman project moving average
self.CP2 = self.p.periodSlow * ((self.movavSlow(0) - self.data.close(-1)) / self.p.periodSlow) - ((self.movavFast(0) - self.data.close(-1)) / self.p.periodFast)
def next(self):
# Market Direction Indicator
MDI = 100 * (self.CP2[-2] - self.CP2[-1]) / ((self.data.close[-1] + self.data.close[-1]) / 2)
# Create lines
self.l.CP2[0] = self.CP2[-1]
self.l.MDI[0] = MDI
print("CP2: ", self.CP2[-1], " MDI: ", MDI)
@kemoduodejuxi said in I seem to have some problems when I trade with CCXT:
self.low_of_last_candle = self.datalow[0]
You are trying to access the datafeed's data in your __init__ method. The datafeed will have no data loaded at this point.
Please see the following docs: https://www.backtrader.com/docu/concepts/#lines-delayed-indexing
You are printing the position right after the buy method is called. The position is not updated at this point - it will get it value only after the buy order will be completed.
I've updated your code with:
def notify_order(self, order):
print('{}: Order ref: {} / Type {} / Status {}'.format(
self.data.datetime.date(0),
order.ref, 'Buy' * order.isbuy() or 'Sell',
order.getstatusname()))
if order.status == order.Completed:
print(self.position)
and the output is now:
Starting Portfolio Value: 60000.00
1999-09-16: Order ref: 1 / Type Buy / Status Submitted
1999-09-16: Order ref: 1 / Type Buy / Status Accepted
1999-09-16: Order ref: 1 / Type Buy / Status Completed
--- Position Begin
- Size: 1478
- Price: 40.53125
- Price orig: 0.0
- Closed: 0
- Opened: 1478
- Adjbase: 40.859375
--- Position End
...
( I was running your code on some yhoo-1996-2015 sample data)
Only LineSeries inherited objects (data feeds) could be added to the Cerebro engine using adddata method.
AFAIU Alpaca API provides the backtrader store AlpacaStore, so you should use the AlpacaStore.getdata method to get the backtrader compatible data feed to be added to Cerebro.
On a different matter:
Please use the backtick or grave accent to wrap you code - this way it will be easier for everybody to read you posts ( see the header on this page )
Please include the backtrace / error info as a text and not images.
@kemoduodejuxi said in I'm interested in this strategy, but there seems to be a problem.:
class a (bt.Indicator):
I believe you should have defined the 'lines' class property for your 'a' indicator class (similarly to the other Indicator inherited classes in the code above)
@chewbacca said in Rollover multiple contracts [help needed]:
File "/home/user/PycharmProjects/backtrader/Strategies/VixLive.py", line 615, in notify_timer
self.initial_position = self.position.size
File "/home/user/PycharmProjects/backtrader/venv/lib/python3.8/site-packages/backtrader/lineseries.py", line 461, in _ _ getattr _ _
return getattr(self.lines, name)
AttributeError: 'Lines_LineSeries_LineIterator_DataAccessor_Strateg' object has no attribute 'position'
The error backtrace above seems strange. Accessing the position property in the strategy should call the bt.Strategy's getposition method:
class Strategy:
...
def getposition(self, data=None, broker=None):
data = data if data is not None else self.datas[0]
broker = broker or self.broker
return broker.getposition(data)
position = property(getposition)
This should be reflected in the error backtrace. It looks like the self.position was called from the indicator and not from the strategy.
It could be useful to if you could share the whole VixLive.py code?
@Jonny8 said in Init data gets not recalculated in optstrategy:
pnl = cerebro.broker.get_value() - globalparams["cash"]
Just a note:
I'm not sure it is a good idea to query the info from the original Cerebro instance's broker.
In multi-CPU case the original Cerebro instance's broker is not updated. In single CPU case (as in the code above), although the same broker instance is used for all the optimization parameter permutation runs, the broker state is reset for each permutation. So querying the broker in this case will return the value for the last optimization run.
It is better to rely only on the data gathered by the analyzers.
Please use the new invite link ( I've set it to not expire ): https://discord.gg/XENRWJzZ36
I've created a discord server. Here the invite: https://discord.gg/EwpKMUbd
Hope it will be useful for the community.
It would be helpful to share your code.
I guess you just need to use the absolute path to your file (even if it is on Desktop it has an absolute path - you can see it using right-click->Properties->Location)
@dark_light said in how to add a big data (over limit the memory) to Backtrader(batch or else?):
32G memory,but minitus data is 200G
how many instruments are you going to test simultaneously?
AFAIU 1 min data for a single instrument per year will take ~5 MB in memory ( 20B per bar * 60 (min per hour) * 16 (hours in ext session) * 252 (trading days in year)).
@Rob-Shih said in Is it possible to run a strategy starting with a certain amount of the stock already bought?:
is it possible to run a strategy where it starts out with a portion of your cash already bought into the stock?
There are some scenarios where back-testing should continue from some point in time - where the back-testing results for the earlier period of time are already available ( stored somewhere ). Sometimes both the analyzer data and broker position should be loaded prior to continuing the back-testing.
In my case I'm doing it in my analyzer create_analysis method - but the same could be accomplished in strategy itself.
Something like this:
class MyAnalyzer(bt.Analyzer):
# strategydb is something I'm using to persist the information about
# backtesting results for specified time period - exact details are not
# important right now.
params = (
('strategydb', None), # strategy database is passed as a parameter
)
def create_analysis(self):
if self.p.strategydb:
# we've got a database with potential history data
# use it for calculating incremental optimization results
... some code here to retrieve the previous results from strategydb
# this will update the position directly inside the strategy broker
data_pos = self.strategy.broker.positions[self.data]
# pos_size,pos_price and stop_date are loaded from strategydb - don't ask how
data_pos.set(pos_size, pos_price)
# in some scenarios broker requires the datetime to be specified
data_pos.datetime = datetime.combine(bt.num2date(sparams.stop_date), time())
It may look like cheating - but it works in my case - for back-testing only of cause ;-)
Could you print the order reference number order.ref every time the order info is printed ?
Would be helpful if you could provide the log of your run and the relevant part of the data feed.
As a note: where do you set the starting cash amount ? ( it is 10000 by default - is it enough for you ?)
Thanks for reporting. Issue opened: https://github.com/backtrader2/backtrader-docs/issues/2