@crazy25000 said in progress bar / ETA during optimization:
Just finished implementing
tqdmfor optimizations. The status bars are inline and stay at the top. This way I can view the ETAs and overview of the opt runs:
Excellent!
Do you have a minimal code example?
Hi all,
I have a strategy that requires my next() to be called on every tick of the price. I have succesfully implemented the ccxtbt module which bridges the gap between backtrader and my broker.
This works great, however it doesn't seem to support live ticker data. When I try to use bt.TimeFrame.Ticks I get the following error:
ValueError: backtrader CCXT module doesn't support fetching OHLCV data for time frame Tick, comression 1
The lowest I can go seems to be bt.TimeFrame.Minutes with compression=1.
But in the background I can see the following network traffic to my broker:
2021-02-14 21:19:23.265920 - fetch_ohlcv - Attempt 0
Fetching: ETH/BTC, TF: 1m, Since: 1613337540000, Limit: 30
Request:
GET https://api.binance.com/api/v3/klines?
symbol=ETHBTC&
interval=1m&
limit=30&
startTime=1613337540000&
endTime=1613337563766
{'User-Agent': 'python-requests/2.25.1', 'Accept-Encoding': 'gzip, deflate'} None
Response:
GET https://api.binance.com/api/v3/klines?symbol=ETHBTC&interval=1m&limit=30&startTime=1613337540000&endTime=1613337563766 200 {
'Content-Type': 'application/json;charset=UTF-8',
'Content-Length': '144',
'Connection': 'keep-alive',
'Date': 'Sun, 14 Feb 2021 21:19:24 GMT',
'Server':'nginx', 'x-mbx-uuid': '2e6s5db1-c4b4-49b2-b476-5358206245f2',
'x-mbx-used-weight':'12',
'x-mbx-used-weight-1m': '12',
'Strict-Transport-Security': 'max-age=31536000; includeSubdomains',
'X-Frame-Options': 'SAMEORIGIN',
'X-Xss-Protection': '1; mode=block',
'X-Content-Type-Options': 'nosniff',
'Content-Security-Policy': "default-src 'self'",
'X-Content-Security-Policy': "default-src 'self'",
'X-WebKit-CSP': "default-src 'self'",
'Cache-Control': 'no-cache, no-store, must-revalidate',
'Pragma': 'no-cache',
'Expires': '0',
'Access-Control-Allow-Origin': '*',
'Access-Control-Allow-Methods': 'GET, HEAD, OPTIONS',
'X-Cache': 'Miss from cloudfront',
'Via': '1.1 fc8f1559s65d4fs6dd42c7d90.cloudfront.net (CloudFront)',
'X-Amz-Cf-Pop': 'AMS50-C1',
'X-Amz-Cf-Id': 'eQ965__94s8fd4s6fs55sd0-8ab61s46sd6sYx-d15sd6s6d5hVg=='} [
[1613337540000,"0.03718400","0.03720500","0.03718300","0.03719000","36.00300000",1613337599999,"1.33899282",85,"11.29600000","0.42016946","0"]]
2021-02-14 21:19:24.014237 - fetch_ohlcv - Attempt 0
Fetching: ETH/BTC, TF: 1m, Since: 1613337540000, Limit: 30
Request:
GET https://api.binance.com/api/v3/klines?
symbol=ETHBTC&
interval=1m&
limit=30&
startTime=1613337540000&
endTime=1613337564524
{'User-Agent': 'python-requests/2.25.1', 'Accept-Encoding': 'gzip, deflate'} None
Response:
GET https://api.binance.com/api/v3/klines?symbol=ETHBTC&interval=1m&limit=30&startTime=1613337540000&endTime=1613337564524 200 {
'Content-Type': 'application/json;charset=UTF-8',
'Content-Length': '144',
'Connection': 'keep-alive',
'Date': 'Sun, 14 Feb 2021 21:19:24 GMT',
'Server': 'nginx',
'x-mbx-uuid': 'as56d5a6-70cc-4967-bsd51-1a21sd8545e',
'x-mbx-used-weight': '13',
'x-mbx-used-weight-1m': '13',
'Strict-Transport-Security': 'max-age=31536000; includeSubdomains',
'X-Frame-Options': 'SAMEORIGIN',
'X-Xss-Protection': '1; mode=block',
'X-Content-Type-Options': 'nosniff',
'Content-Security-Policy': "default-src 'self'",
'X-Content-Security-Policy': "default-src 'self'",
'X-WebKit-CSP': "default-src 'self'",
'Cache-Control': 'no-cache, no-store, must-revalidate',
'Pragma': 'no-cache',
'Expires': '0',
'Access-Control-Allow-Origin': '*',
'Access-Control-Allow-Methods': 'GET, HEAD, OPTIONS',
'X-Cache': 'Miss from cloudfront',
'Via': '1.1 fc8f1s84d64fsd54fs5d5sdsdfsdcx56590.cloudfront.net (CloudFront)',
'X-Amz-Cf-Pop': 'AMS50-C1',
'X-Amz-Cf-Id': 'vV6s4d8g4hgfh5f46sdsfcAZ-3W6k5MTn__GtwA0-m9l7Rbw=='} [
[1613337540000,"0.03718400","0.03720500","0.03718300","0.03719100","36.41400000",1613337599999,"1.35427825",87,"11.64200000","0.43303754","0"]]
As you can see (in the latest line of the responses) the ccxtbt module does in fact continously poll my broker for the latest OHLCV data, and receives the latest tick price as the "Close".
So now my question is: How can I make this latest tick price be my dataClose[0], and call the next() of my strategy?
Thanks
Hi all,
I've succesfully implemented ccxtbt to backtrader and can trade live on Binance. However there are some things unclear to me:
Looking at de debug output I can see ccxtbt is continously fetching the newest ohlcv ticks, multiple times per second.
Using bt.Timeframe.Minutes my next() is only called once per minute. However I want my next() to be called every new tick. When I apply the bt.Timeframe.Ticks as supported by Backtrader, I get the following error:
ValueError: backtrader CCXT module doesn't support fetching OHLCV data for time frame Tick, comression 1
Is this correct? Is there no way for me to get the individual tick price live, and feed it to my next() ?
Apparantly I had simply forgotten to add the strategy when live trading using
cerebro.addstrategy( TestStrategy)
This also meant none of the indicators were initialized...
Hope this helps someone!
Hi all,
I'm starting to try live trading with my startegy which has up to now only been backtested.
I have a custom indicator and a custom analyzer which both run fine during backtesting, however now I'm trying to run the same strat live (using the ccxtbt store for Binance) and this is what happens:
class customIndicator(bt.Indicator):
lines = ('someLine')
params = (('dataClose', 0))
plotlines = dict(someLine = dict(_name='someLine', color='green', linewidth = 1, alpha=0.9))
def __init__(self):
self.l.someLine = ...
class customAnalyzer(Analyzer):
def __init__(self):
self.var = 0
def start(self):
pass
def next(self):
if (self.strategy.customIndicator.l.someLine[0] > 10):
self.var += 1
...
def stop(self):
...
def get_analysis(self):
return dict(var = self.var)
Obviously I simplified the example code to it's essence.
This is the error I get:
File "...main.py", line 221, in next
if (self.strategy.customIndicator.l.someLine[0] > 10):
File "...\lib\site-packages\backtrader\lineseries.py", line 461, in __getattr__
return getattr(self.lines, name)
AttributeError: 'Lines_LineSeries_LineIterator_DataAccessor_Strateg' object has no attribute 'customIndicator'
So for some reason the customIndicator has not yet been initialised, but the
def next(self):
In the analyzer is already being called. Is this normal? Why is this different during Live trading than backtesting?
Kind regards,
Fred
Now the Bokeh plotting works again, however it seems all of the parameters given in the object constructor are ignored:
b = Bokeh( filename='FREBT.html', # Bokeh plotting object constructor
title='FREBT',
style='bar', # Candlesticks
plot_mode="single", # single / tabs
plot_width=1550, #
plot_height_data = 500, # See Scheme.py on Github "backtrader_plotting " for all possible params concerning Bokeh layout
plot_height_observer = 120,
plot_height_indicator = 100,
legend_click = "hide", # Hides line clicked on. "Mute" disables this
legend_background_transparency = 0.9,
tools = "pan,wheel_zoom,box_zoom,reset,save")
cerebro.plot(b)
Any ideas?
@richard-klein That fixed it! Well done!
What could be the cause of this?
Hello all,
I had a working program running in my pipenv shell untill I updated all dependencies using
pipenv update
Now I get the following error when my code tries to plot the result of a backtest:
Traceback (most recent call last):
File "main.py", line 942, in <module>
cerebro.plot(b)
# Plot Bokeh object
File "C:\Users\Craptop\.virtualenvs\FREBT-VbURPJvn\lib\site-packages\backtrader\cerebro.py", line 991, in plot
start=start, end=end, use=use)
File "C:\Users\Craptop\.virtualenvs\FREBT-VbURPJvn\lib\site-packages\backtrader_plotting\bokeh\bokeh.py", line 511, in plot
df: pd.DataFrame = self.build_strategy_data(obj, start, end)
File "C:\Users\Craptop\.virtualenvs\FREBT-VbURPJvn\lib\site-packages\backtrader_plotting\bokeh\bokeh.py", line 473, in build_strategy_data
for lineidx in range(obj.size()):
AttributeError: 'LinesOperation' object has no attribute 'size'
My entire code still works fine, it's just the plotting (I'm using Bokeh) that seems broken. I have been searching for a solution but can't find one. Can anyone point me in the right direction?
Regards,
Frederik
Fixed by using
self.strategy.indicatorName.lines.lineName[0]
Should've read the documentation more closely...
Hi all,
I'm creating a custom analyzer in which I need to compare two lines from a custom indicator. This custom indicator has already been added to my strategy.
Do I need to instantiate the indicator again in the __init__ of my analyzer? Is this not inefficient? It is a large indicator with lots of lines.
Can I somehow "import" or "reach" the current line value of an indicator in my strat each next in my analyzer?
I hope this explanation is a bit clear :)
Thanks