Why MATLAB and not Python, R, Tensorflow, or similar open and free software?

Why does ANDATA primarily use MATLAB® and not free and open software like Python, R, Tensorflow, ROS, etc. in the professional projects?

There are several reasons, why ANDATA primarily uses MATLAB® and not open and free software in their professional projects, like Python, R, Tensorflow, ROS, etc.:

• Open source may be free of licenses but is not free of cost and effort. Efforts and costs for configuration and maintenance of the development environments should not be underestimated. As professional company also these efforts and the overall costs for the final solutions have to be taken into account.
• Not with all open source software a sustainable development and maintenance is guaranteed. Professional software vendors deliver a much stronger commitment and obligation. This is the base for much more sustainability.
• MATLAB is very competitive with respect to performance. For example, check
  • https://de.mathworks.com/discovery/matlab-vs-r.html
  • https://de.mathworks.com/products/matlab/matlab-vs-python.html
• MATLAB delivers a universal, integrated development environment with tons of concerted algorithms and functions.
• MATLAB delivers a much easier and more configurable possibilities for graphical display for information, data, and results.
• Because ANDATA is primarily using MATLAB there is no conditional need that ANDATA's customers must also use MATLAB. The final solution can most of the time be implemented independently.
• Some major industrial customers already refuse the use of open source in professional projects because of legal issues, especially in safety critical domains.

Of course this page should not be an advertisement pamphlet for MathWorks®. Neither should it be a statement against open source, which we highly respect. Same arguments may be true as well for Mathematica, Maple, or similar others!

As well as we are benchmarking, testing, and comparing different Machine Learning, simulation, and technical computation programs all the time, we are doing such with development environments and scientific software packages. In various professional, research and academic projects we are using Python, Java and other environments, respectively we are combining and integrating these in MATLAB and vice versa (e.g. check https://de.mathworks.com/help/matlab/matlab-engine-for-python.html). In addition we integrate Machine Learning models form Python in Brainer.

Nevertheless our above opinions are confirmed most of the time, if modelling and simulation is the core business.