Writting an algorithm in C++

[sef88]

For low frequency strategy, it makes more sense to use python.

 

[zenlot]

I always use node js if I can. Benchmarks show that it performs much better than python and C#/C++ in situations with large amounts of data and with a large number of operations. Your four conditions calls for a few simple lines of code anyways. To cover all bases, you would also need to consider conditions for exits, profit targets and time decay.

Do you happen to have links to those benchmarks?

 

[Russell Shuffle]

gonna throw my two cents in here for python if all you're doing in local modeling/forecasting. Super easy and it's plenty fast (I use it on a 5000 stock universe every weekend...)

Google "python rsi". You'll find a Stack Overflow solution that does exactly what you're looking for.

 

[d08]

I always use node js if I can. Benchmarks show that it performs much better than python and C#/C++ in situations with large amounts of data and with a large number of operations. Your four conditions calls for a few simple lines of code anyways. To cover all bases, you would also need to consider conditions for exits, profit targets and time decay.

nodeJS is written in C/C++ so that's impossible. Most of the benchmarks are faulty to begin with. When done properly, C/C++ is magnitudes faster. Python is slower but no-one really uses native python for any larger datasets, modules are used which implement C/C++ underneath.

 

[931]

You can also backtest this prior to any real money on the line.

wrbtrader

Amen.

 

[windwave]

Definitely possible and if you are familiar with c++ then why not stick with it.

 

[MrDickeyFuller]

There is just no reason to not use NumPy. There is no point in comparing "python vs C++". Your implementation in C++ is not going to be faster than NumPy.

That doesn't even touch on the ecosystem you are then getting for free.

 

[igr]

I would like to develop an algorithm in C++ that would analyses stocks price and volume behavior in a single time frame (from 9:30am to 9:33am for example), with a pre-market volume of 50K for the last 250 days and in case of RSI 70, short the position or RSI 30 then buy the position. Is that even possible?

Would it be a faster way to learn whether your idea works to sign up on TradingView and write/backtest a strategy using their Python-like language?

 

[zenmaster]

There is just no reason to not use NumPy. There is no point in comparing "python vs C++". Your implementation in C++ is not going to be faster than NumPy.

That doesn't even touch on the ecosystem you are then getting for free.

Is it possible to develop in Python and execute in C++? The platform I use is developed in C++ and I'm not sure I can build in Python and execute in C++.

 

[igr]

Is it possible to develop in Python and execute in C++? The platform I use is developed in C++ and I'm not sure I can build in Python and execute in C++.

Yes. For example, Tensorlow works like that.