If you do data analysis, toss about all kinds of data drawing (such as regression, interpolation, numerical calculus, etc. ), python is obviously very useful, all kinds of expansion packages are fast and extensible (such as Numpy, SciPy, Pandas), and the written programs are short and pithy, easy to modify and maintain.
If you want to build a mathematical model, you'd better use a linear model, and then use the CPLEX optimizer, which is the CPLEX mentioned by @ Cui Youzhi.
It supports multiple languages, including C, C++, Java and MATLAB. You only need to be familiar with one of these languages, so you can write model files and solve them in a relatively short time. If you are interested in these.
I'm not familiar with it, so there are two other softwares GAMS/AMPL. They seem to embed many efficient solvers in it, and then express the model with unified and relatively direct grammar rules, but the shortcomings may be necessary.
To get a license, the free version can only solve models within a certain scale, and can also solve some relatively simple nonlinear models. But if you write a nonlinear model, and the scale is still large or
It's complicated, so bless you, you need to think of something more fancy. . . I don't know what software is suitable.
If you want to perform symbolic operations, such as solving symbolic calculus and finding the analytical solution of differential equations, Wolfram Mathematica must be the most applicable, and its grammatical rules are also very direct (its goal is natural language). This software corresponds to a website Wolfram|Alpha, which can run a single line of code. I often use it to do some single-player operations. I can try it. At the beginning, my roommate and I participated in the beauty contest. He wrote the program in mathematica. The effect is very good, considering our needs at that time, we really need to write in mathematica.
Until/very
In matlab, the above points can be basically achieved, but my personal experience tells me that in mathematical modeling, MATLAB is "broad and not very professional", if it is only relatively simple but needs a variety of analytical solutions.
Legal issues are very useful. Of course, if it involves large-scale matrix operations, or you can use some of your own toolbox, then matlab is still very powerful (experience the powerful feeling of matlab)
It takes a lot of time), or, you prepare some code files for some specific problems or specific algorithms in advance. (This paragraph is purely a personal opinion. )
But in any case, I use MATLAB and Cplex most often in scientific research, and occasionally I use python. Learn what you need, mathematical modeling is not the purpose, but solving problems is the purpose.