Python + Qt for heavy scientific calculations and graphics - a myth?

Question

Python

Hello,
I am a complete noob in programming, but I have basic ideas.
Now I am in the role of a software development customer.
They write software for visualization and processing of curves of real data of the type (Y,t) with a number of points of ±500 thousand.
All this needs to be zoomed in, scaled, rebuilt, quickly twisted, turned on one curve, turned off another, displayed grids of numerical calculations in dynamics and now everything is terribly slow.
The developers chose the Python + Qt4 bundle, although I felt in my gut and said, write in C ++.
I screwed up and all this cannot be avoided with such an architecture at the core?

Answer 1

In the technical task, indicate the desired characteristics of the smooth operation of the interface and ask them to write an example of work with a large amount of data
Most of the scientific python is already written in C - there should be no
problems and there will be a cool and smooth interface, but in ten years, until they write everything they need on the pluses

Answer 2

On a bunch of Python + Qt, you can write something that will visualize data and pretty quickly. I have development experience for visualization of images consisting of 500k points and more taken by zmq. For this, I used the pyqtgraph library . For images, mean calculations, correlations, and so on were performed.
If processing performance is not enough somewhere, then this section can be completely rewritten in C ++, and then called from python.
PS Of course, we must not forget that a lot depends on the direct hands of the developers.

Answer 3

well, in general, I found on the Internet an ordinary program that displays such graphs, well, just in years ...
but on Python, my dears - everything slows down ...

Answer 4

There is nothing better for scientific data processing than OriginPro from OriginLab. Its free counterpart is qtiplot.