matplotlib VS feedgnuplot

By using pybind11 that one can be written with far fewer lines of code.

Thanks for the post. The obvious comparison is feedgnuplot: https://github.com/dkogan/feedgnuplot/

That works similarly in that it plots standard input. The backend is gnuplot, which is a double-edged sword: it's far more full-featured than brplot, but almost certainly is much slower also. I'll try out brplot to see if it would be a good replacement for cases where speed is important. Thanks!

Oh hey Dima.

Feedgnuplot is really slick.

https://github.com/dkogan/feedgnuplot

It's in the debian repos too.

Is there a declarative language or framework to create ad-hoc GUIs that consume structured data from stdin stream and spit-out a GUI?

Like feedgnuplot [1] but not only restricted to graphs.

[1] https://github.com/dkogan/feedgnuplot

I found using shell as an interactive environment to be pretty productive using https://github.com/dkogan/feedgnuplot and https://github.com/dkogan/vnlog. The filesystem becomes your state (instead of in memory state of your Python interpreter) which forces you to write Unix-style tools. Plotting with feedgnuplot spins up an interactive Qt plotter which I often used to explore 3D plots. It's not "inline" and fancy and does take a bit of grokking but I eventually found it more productive than Jupyter, especially as my development moved away from Python.

One of the tools I absolutely love is feedgnuplot which presents a stdin CLI interface to gnuplot.

Also, if one doesn't want to learn Gnuplot's DSL try using feedgnuplot which presents a stdin interface for whitespace delimited tabular data.

Write space delimited tabular data (ideally in vnlog format) and plot it using feedgnuplot. Also helps decouple concerns (data generating application focuses on generating data).