A tool for developing quantitative strategy model

github.com

3 pointsJudyrabbit4 comments

> Look at the Python code first:

close=[12.99,12.72,…] s = pd.Series(close) ss = s.shift() ma = (s-ss).fillna(0).max()

SPL code: =A1.(if(#==1,0,~-~[-1])).max()

Well, this level of cryptic stuff is nightmare for debugging and maintenance. So i think i prefer Python. If loops are slow, perhaps Cython or C/C++/Rust interop will get it done. Or there is always Julia for quant strat professionals who recoil in horror at the sight of C/C++/Rust.

For coding and maintenance, compared to SPL, Python is THE nightmare（~-~）. Python provides so many types of set data, each has its own adaptation scenarios and operation rules, which is dizzying. And the transitions between them are particularly smooth, and if you're not careful, you'll calculate out something you don’t recognize. A a result, programming basically relies on googling, and you don't know why even you are right, and you’ll have to google again next time. SPL has just one type of set data, which can fit in any case.

0. Python is not that bad!

1. Why not just use Rust or Julia?

Rust is not nearly as concise as python, it is worse nightmare than python. Python is not good enough, but compared to these, it is the best (except that the execution is slower). However , SPL is better than Python.

Reply on news.ycombinator.com