Hi,
I still trying to optimize my simulation runs.
Optstrategy is good but I still have the issue with memory leaks and if it crashes all results are lost, so could split up and let run multiple of them but still...
What I was thinking was to create a Broker, that handles multiple other Broker, so that at each next a strategy with it's own account (broker) is called.
I don't got enough inside into the interplay of cerebro and broker if that would be an option.
@kjiessar
Aaand again. After long debugging session I found the culprit and it wasn't even in the code I had shown.
Errorness:
self.sell(data=self.data0, exectype=bt.Order.Limit, price=self.datas[0].close, size=entry["size"], valid=datetime.timedelta(minutes=self.p.tis))
correct one:
self.sell(data=self.data0, exectype=bt.Order.Limit, price=self.datas[0].close[0], size=entry["size"], valid=datetime.timedelta(minutes=self.p.tis))
Note the proce parameter and the last [0]
And more Information:
GitHub: BT Transaction Analyzer
def next(self):
# super(Transactions, self).next() # let dtkey update
entries = []
for i, dname in self._idnames:
pos = self._positions.get(dname, None)
if pos is not None:
size, price = pos.size, pos.price
if size:
entries.append([size, price, i, dname, -size * price])
if entries:
self.rets[self.strategy.datetime.datetime()] = entries
self._positions.clear()
@kjiessar
In the above code sample a run is performed, just tried to put in the relevant information
Some more info not sure if it can help but thought couldn't hurt
Trade:
1970-01-01 19:45:46,-5.0,2.224e-05,0,,0.0001112
1970-01-01 19:45:54,-5.0,<backtrader.linebuffer.LineBuffer object at 0x0000029F82F3DF70>,0,,0.00011485
OHLC
1970-01-01 19:45:53+00:00,1629356760,2.271e-05,2.271e-05,2.271e-05,2.271e-05,0.0,0
1970-01-01 19:45:54+00:00,1629356820,2.279e-05,2.297e-05,2.279e-05,2.297e-05,0.0,0
1970-01-01 19:45:55+00:00,1629356880,2.275e-05,2.275e-05,2.275e-05,2.275e-05,0.0,0
@frenzyy
Alternative
https://www.backtrader.com/docu/strategy/#how-to-buysellclose
take a look at the parameters
price, limit, plimit and exectype
Hi,
the output of the analyser gives object str instead of the actual price.
Any idea what I'm doing wrong here?
cerebro.addanalyzer(bt.analyzers.Transactions, _name='transactions')
d =strat.analyzers.getbyname("transactions").get_analysis()
print(d)
One correct and one outputted wrong
OrderedDict(
[(datetime.datetime(1970, 1, 1, 19, 9, 27),
[[23.10281162, 2.05e-05, 0, '', -0.00047360763821]]),
[...]
[[-23.10281162, <backtrader.linebuffer.LineBuffer object at 0x000001847D964400>, 0, '', 0.0005489228040912]]),
[...]])
@kjiessar
Its a bit embarassing to say but I found the issue
data = store.getdata(dataname=symbol, name=symbol,
timeframe=bt.TimeFrame.Minutes,
compression=1, ohlcv_limit=70 )
#fromDate is missing
I had the parameter but with the wrong name, becuase of all the dynamic handling of the parameters, no error was raised.
Sorry for the trouble.
Hi,
I'm using ccxtbt or my own fork of it ccxtbt-fork.
My issue is, that my strategy assumes time slices of a minute and the strategy produces slices of around 10 seconds (could be unrelated and based on runtime [debugging therfore slow]).
data = store.getdata(dataname=symbol, name=symbol,
timeframe=bt.TimeFrame.Minutes,
compression=1, ohlcv_limit=70 )
Thats how i create the feed. I changed code in the broker (fixed bugs with order handling). But don't think I changed anything regarding the handling of the time.
I just reinstalled the original version to check if I messed up but couldn't find any difference.
So my expected behaviour is, that my Strategy is called once a Minute.
Am I expecting somthing on wrong assumption? ( Don't think so, why choose timeframe and compression if not used)
Any idea where to start to look? Is that a bug or just missing feature not yet implemented in bt-ccxt-store?
@kjiessar
Aaaaand 5 minutes later. Trying to fix it, it seems that the issue is something else. Will come back as soon as I found out
Hi,
I know it seems stupid but can happen, that more tan one transaction is triggering at a time.
The bt.analyzers.Transactions will use a map and use the date as key putting in only 1 of the transactions, overriding the beforehand added.
Code Link
def next(self):
[...]
if entries:
self.rets[self.strategy.datetime.datetime()] = entries
This can compromise your analysis, just be aware of it.
Documentation issue:
Custom Schemes
class CommInfo_Fut_Perc_Mult(bt.CommInfoBase):
params = (
('stocklike', False), # Futures
('commtype', bt.CommInfoBase.COMM_PERC), # Apply % Commission
('percabs', True), # pass perc as 0.xx
)
comminfo = CommInfo_Fut_Perc_Mult(
commission=0.001, # 0.1%
mult=10,
margin=2000 # Margin is needed for futures-like instruments
)
cerebro.addcommissioninfo(comminfo)
last line should be:
cerebro.broker.addcommissioninfo(comminfo)
Hi there,
at the end of a run I'm usually take cerebro.broker.get_value() to sort my strategy paramters.
Now I'm givin OptStrategy a try and can't find the value for each strategy.
If not any preferred way to implement it? E.g. an Analyzer seems a bit overkill.
@kjiessar
Ok tested a bit around.
New assumption: Stupid me forgot to include the commision fees.
That meens i got enoug money for the share to buy but as the fees will be added to the price, I will not get as much as I ordered, because it is downsized as much as needed, to pay the fee.
Correct?
Hi there,
I'm struggling a bit withthe Order.Margin. So here are is my base assumption:
Order.Margin is the status, when:
remaining_cash - all( open_orders size * prize) < 0
Some of my cash is "blocked" because i have open orders and the "free" cash is not enough to "block" enough cash for the new order.
If thats correct, is there an api call that gives me the remaining "free" cash aka margin(?)?
Done:
https://github.com/backtrader2/backtrader/issues/65
Not that good at filling those, so hope it is enough like that.
@borutf
Would agree, looks like a bug:
if self.p.todate is not None:
period2 = (self.p.todate.date() - posix).total_seconds()
urlargs.append('period2={}'.format(int(period2)))
if self.p.todate is not None:
period1 = (self.p.fromdate.date() - posix).total_seconds()
urlargs.append('period1={}'.format(int(period1)))
Hi everybody,
sorry if this is noob question but I'm struggeling with getting the value (is that called portfolio value?) after a run.
I'm currently using cerebro.broker.get_value() as I can also see it my plots
The small values are because cryptotrading data is used.
So my issue here is:
I'm starting with 0.0019 BTC here and buy some other crypto.
Even if this other crypto drops to nearly 0, lets just say the remaing complete value of my portfolio is 0.000000001BTC.
If I sell it (so adding the remaining value), it should still have remaining cash +0.000000001BTC
It seems that I'm still missinterpreting something here.
Another topic: why is my plot so damn ugly compared to the on in the docs? (can create another thread if it isn't a short answer)
@vladisld
Another thing I forgot ( sorry it's hard to find the relevant parts here) is that I'm saving the plots:
def saveplots(strats, numfigs=1, iplot=True, start=None, end=None,
width=640, height=360, dpi=300, use=None, file_path = '', **kwargs):
plotter = plot.Plot(**kwargs)
figs = []
for stratlist in strats:
for si, strat in enumerate(stratlist):
rfig = plotter.plot(strat, figid=si * 100,
numfigs=numfigs, iplot=iplot,
start=start, end=end, use=use, width=width, height=height, dpi=dpi, constrained_layout=True, )
figs.append(rfig)
for fig in figs:
for f in fig:
f.savefig(file_path, dpi=dpi, width=width, height=height, bbox_inches='tight', )
I tried a lot like f.clf() and so on but nothing seemed to help.
@vladisld
Hi sorry for not providing that:
Mhh maybe not the only issue I'm facing. Still continiously growing.
My current workaround:
results = []
with Pool(multiprocessing.cpu_count, maxtasksperchild=1) as pool:
result =pool.apply_async(run)
results.append(result)
[result.wait() for result in results]
pool.close()
pool.join()