@ard9000 said in Adding data after cerebro.run():
Adding data after cerebro.run()
Not after.
@ard9000 said in Adding data after cerebro.run():
Can .adddata() be called during .run() ?
Not during.
Any other suggestions?
Yes.
@ard9000 said in Adding data after cerebro.run():
our FIX interface we developed for our firm with BackTrader.
You are a firm, pay to get something done which fits your needs instead of being cheap.
@Simone said in Can somebody explain me how to use the heikin ashi indicator into a strategy:
You say you want to work with the INDICATOR.
@backtrader said in Can somebody explain me how to use the heikin ashi indicator into a strategy:
The indicator version has with 4 output lines
ha_open
You get information for the INDICATOR
@Simone said in Can somebody explain me how to use the heikin ashi indicator into a strategy:
but is using hk.close instead of ha_close that you told me.
Apparently you DON'T want the INDICATOR. You want a standard data stream transformed into a Heikin Ashi data stream.
But first, if you have the indicator ... with a line named ha_close, you don't do this
@Simone said in Can somebody explain me how to use the heikin ashi indicator into a strategy:
self.sma=bt.indicators.SimpleMovingAverage(self.hk.ha_close[0], period=self.params.speriod)
Because as pointed out by @momentum, you are passing the CURRENT value (which at the moment you actually create it ... it doesn't exist) to the SimpleMovingAverage
You pass the ENTIRE line
self.sma=bt.indicators.SimpleMovingAverage(self.hk.ha_close, period=self.params.speriod)
If you want the transformed data you add a filter
data.addfilter(bt.filters.HeikinAshi)
Please read this post
Let me venture into the crystall ball territory ... you did
pip install oandapy
(Yes I know, my crystal ball is incredible ... but it doesn't work with lottery numbers unfortunately)
The documentation explicitly states (Docs - Oanda)
pip install git+https://github.com/oanda/oandapy.git
Hopefully you see the difference. Furthermore, you are installing something for Oanda REST v20 which isn't supported by the built-in functionality.
That's why another part of the website lists 3rd party resources such as
@elihubogan said in Backtrader, oandapy and OandaError issues:
can't make much progress in the QuickStart if the module won't load
The Quickstart doesn't reference Oanda.
@edge7 said in Backtest accuracy:
Am I right?
Of course. Targeting 5-min bars is going to give a lot more accuracy than targeting 1-day bars.
@Darshil-Shah said in Cerebro.run() Type Error:
I am new to this community and have just started using backtrader.
Basics
As such (and the top of each and every page in the forum)
For code/output blocks: Use ``` (aka backtick or grave accent) in a single line before and after the block. See: http://commonmark.org/help/
Without that your code is basically unreadable for human beings with two eyes (those with three may possible add an extra visual dimension and put the code back together)
There are a couple of obvious things in any case:
@Darshil-Shah said in Cerebro.run() Type Error:
It shows TypeError: must be real number, not LineBuffer
Haven't you considered that sharing the error (the trace) is as important as the code?
This is telling you, that you are passing the line to an assigment, where you should be passing a float (or something which can be converted directly to a float like an int
@Darshil-Shah said in Cerebro.run() Type Error:
if date in date_sentiment: self.sentiment = date_sentiment[date]
That code at the beginning is obviously doomed to fail, because there is no date_sentiment defined anywhere even if the code is unreadable.
@Darshil-Shah said in Cerebro.run() Type Error:
self.date = self.data.datetime date = bt.num2date(self.date[0]).date()
A lot easier (and timezone aware if timezones are used) if you do:
date = self.data.datetime.date()
Sincerely: you need to make up your mind as to what you want. What you post makes no sense. You probably have an idea about the direction in which you want to go.
If your take-profit/stop-loss is based on the value increase/decrease of an asset you have acquired, record the initial value and exit when the value has gone beyond your limits.
I would argue your problem is in how you call the indicator. But it is a long shot, given that part isn't shown.
You can always have your own signals as an extra input, overlay any number of indicators, run the backtrader script once a day and let it act accordingly.
Yes.
But it will of course take development time from your side.
@jabbarabdullah said in Assigning independent value for multi assets:
data_d_cash = portfolio_cash * (1/len(datalist)).
Apply then that in your loop.
The indicator version has with 4 output lines
It works exactly as any other indicator. What is the problem?
A reference is always a good starting point. It does let people be on common ground. For example
You probably want to read the documentation about line delaying and stage-1, stage-2 operations.
What some people would call a "canonical" implementation can be found here ... (which you probably can better grasp after seeing the Stage-1 explanation about operators)
https://www.backtrader.com/blog/2019-05-20-momentum-strategy/momentum-strategy/ and in several community threads
@kr said in Bracket Orders Behavior:
It would be an awesome feature if we could give backtrader finer granularity data just for it to use in the background for smart execution of bracket orders, and give it a main data (daily in my case) to operate as it does right now, i.e. call next() with it, etc.
Incredible as it may seem ... that awesome functionality is built-in from day 1.
Indeed. You seem to be extremely new to indicators, timeframes, periods and resampling. ta-lib has a problem with the implementation of indicators with an "unstable" period, because of a cumbersome implementation, which is easily solved (industry de-facto standard) by using a seed value.
An "Exponential Moving Average" is an example, because the definition of that average is that of a first order infinite impulse response filter. All values are always taken into account which leads to a recursive definition for updating values in real-time.
Please read: https://en.wikipedia.org/wiki/Moving_average#Exponential_moving_average
You can see that the problem is a ta-lib well-known issue: https://github.com/mrjbq7/ta-lib/issues/114
Back to square one: a good reason not to use ta-lib
Your problem is the same problem many other have: you want to do something with daily data which cannot be done with daily data.
There is no custom coding which can help you. You simply need more granularity.
@jabbarabdullah said in Multi Assets Multi Order Executions Problem:
Appreciate if you could explain further on what do you mean by 'NO'.
You say the orders are getting executed twice and the answer is: NO
@ab_trader said in Multi Assets Multi Order Executions Problem:
You have two data feeds, you cycle thru them twice and print same info twice. Check if data name is same as order data name, then print.
On the other side you don't need to go thru all data feeds. Just print order info.
I quoted the code in notify_order where you do exactly what @ab_trader is telling you. You print the execution twice (which is not the same as two executions), because you loop through the datas in notify_order
You can also shoot yourself in the knee and then see if it hurts. The problem is the data provider and the amount of symbols requested from it. Blocking a thread 15 seconds to try to overcome a structural problem is not the way.
Feel free to do it yourself.
The quandl data feed was done for the WIKI data (which is actually no longer being updated). Your best bet is to download the data to a "csv" file and load it from there.
Orders are executed (and evaluated for admission simulating pseudo-execution) on a FIFO basis, just like in a regular exchange.
You have to properly schedule the order. If your sell orders are sent before the buy orders and are Market orders, they will ALWAYS be executed before the buy orders.
@kr said in Bracket Orders Behavior:
Do bracket orders only operate on close prices?
As pointed out by @Robin-Dhillon, you choose the order types (although only a few make actually sense, and what gets accepted is for sure broker dependent).
The general answer is no. Order execution depends on the order type.
@kr said in Bracket Orders Behavior:
If not, how does it know if low (stop price) or high (take profit) was hit first?
It doesn't know and it cannot know it. Orders, just like in real exchanges are executing on a FIFO basis.
@kr said in Bracket Orders Behavior:
Will it help if we give backtrader smaller timeframe data and resample to larger timeframe?
It's not about helping backtrader. You can only help yourself by using the smallest possible timeframe for execution. Anyone letting you provide help to decide which order is executed first is giving you something not realistic and biased.