The subject of visual programming is quite old; however, I think it's worth revisiting. I agree with one of the previous threads that points out that visual programming is a failed approach, with one exception, however.
This exception relates to describing data flows when one backtests a strategy on several portfolios of instruments. While this is not visual programming in its original sense,
visualization is an approach that has its merits and beats other ways of managing data.
I built my first trading algorithms in Java, C, but I finally settled with R (using quanstrat and related packages). Recently, when I started testing my strategies on multiple
portfolios of stocks, I was inundated with data so I had to build a visual environment that aims to simplify the workflow through automated parallel processing.
You still have to write code, but life gets easier when you structure your code around data flows that are described visually. This is especially important if you want to use
data from different sources in the actual production stage.
I am sharing my software in exchange for feedback so I can improve. The environment is still in development, but the code block formats have been settled, or at least it seems they have been.
Please, have a look at and let me know what you think. I really appreciate any input. (Don't forget to download examples in the download section).
The software can run scripts written in R (tested and it works), Python (not tested yet, but should work), and compiled code (works if one builds it according to specifications).
Thank you very much in advance.
PS
here's a sample screenshot (simulation of parallel data processing):
and another one (this is me just stress-testing the system):
and another one (getting data from Yahoo finance):