Is MATLAB being phased out by Python?

It will not be replaced. Simply put, Python is a universal language, which can do everything, while matlab is good at calculation.

Compared with Matlab, Python has the following advantages:

Python is a general programming language. Numpy, scipy and matplotlib, which realize scientific computing functions, are just Python libraries and packages. In addition, Python has libraries and packages for various purposes, such as PyQt and wxPython for GUI and Django and Flask for Web.

Compared with Python, Matlab has the following advantages:

It is specially developed for numerical calculation, and has the largest number of libraries, users and books in the field of numerical calculation.

If you do strategy research and data analysis, the two functions are similar, but you should choose matlab because:

If you still want to do some IT-oriented work such as web crawler and data cleaning, Python is better.

MATLAB is an advanced technical computing language and interactive environment, which is used for algorithm development, data visualization, data analysis and numerical calculation. Using MATLAB can solve technical calculation problems faster than using traditional programming languages (such as C, C++ and Fortran).

With the continuous supplement and improvement of MATLAB toolbox, M language has gradually become a quasi-universal standard language in engineering. Official website called it MATLAB-the language of technical calculation.

The science and engineering majors in general universities have set up MATLAB-related courses, which are optional or compulsory. Many newly published textbooks and computer-aided teaching tools began to choose MATLAB.

MATLAB has gradually gained popularity with its easy-to-learn grammar, friendly interface and perfect document system, and will continue to expand its control territory.

However, MATLAB also has great limitations. First, the price. As commercial software, it is expensive to use genuine license. For example, the single license price of the cheapest student version core component is $99, and if you want to use additional toolboxes, the price of each toolbox is $29. It is conceivable that the business version is more expensive.

Followed by copyright. Mathworks forum has many active users, and the code is valuable, but the copyright belongs to Mathworks company and must be authorized to use it.

Third, it is the perfection of language. There is no doubt about the performance of MATLAB in mathematical calculation, but the actual scientific calculation also includes file operation and interface design. MATLAB is weak or troublesome in these fields. It should be said that MATLAB is not a perfect language.

Also: matlab is widely used in academic circles for simulation, and it is easy to find code reference when doing research;

Grammar is more flexible than python, and matlab basically does not use routines to write programs. The so-called old-fashioned matlab is one word, dry;

There is simulink. Some people say that simulink is useless, but it is actually quite useful, such as communication modeling. In addition, simulink can generate DSP or FPGA code, which is sometimes very useful.

First of all, Python is completely free, and most extension libraries related to scientific computing are also free, and most of them are open source, so the problem of money is completely ignored. There is basically no need to consider copyright issues, and you can use many example programs. Sometimes it needs to be considered, because some authorizations, such as GPL authorization, are "contagious". Consider countries such as the United States, where copyright control is stricter. Many researchers and college students use Python, and many networks provide communication platforms, on which more communication and learning opportunities can be obtained.

Secondly, Python is an object-oriented programming language, which is easier to learn and more rigorous. Python, as a general programming language, has stricter and clearer syntax, and can easily meet high-level requirements such as interfaces, files and packaging. Finally, I have to mention performance. MATLAB, as a scientific computing tool, has experienced almost harsh optimization. What about Python?

To tell the truth, the speed of pure Python is really not so good, but after using Python's scientific computing extension libraries numpy and scipy, the speed is comparable to that of MATLAB.

Another big advantage: open source. You can change the algorithmic details of many scientific calculations.

Portable, Matlab is not as good as Python. But you mainly do research, and the demand in this area should not be high.

Third-party ecology, Matlab is not as good as Python. For example, 3D drawing toolkit, such as GUI, such as more convenient parallelism, using GPU, functionality and so on. In the long run, Python's scientific computing ecology will be better than Matlab. The language is more beautiful. In addition, if there is a certain OOP demand, to build a larger scientific computing system, it is definitely easier to mix directly with Python than with Matlab.

Python, as a general programming language, can be used as a net, a crawler, a script and a gadget.

Most data analysis, image processing, digital signal processing and data visualization can be completely separated from matlab. Especially for enterprise users, the dependence on matlab is not as high as before. Matlab kernel has low efficiency, low execution efficiency and slow simulation speed. Python combined with CUDA can be processed in parallel to speed up the simulation. Not to mention the AI field that is in full swing recently, there is nothing wrong with matlab.

Stop playing matlab. Python is a universal language, and matlab is a charging toolbox. I admit that a toolbox like matlab simulink is very powerful. But learning matlab has no future, because there is a charge, and no company will use it.

In recent years, Python programming language has caused quite a stir in China, and it tends to surpass Java. Originally, this programming language was the hottest in America, and it was widely used. Python is much simpler than Java in terms of overall language difficulty. Especially in the application of operation and maintenance, it is very extensive, so as I said before, in today's era, operation and maintenance will be eliminated sooner or later without learning Python!

But does Python really have such a good employment prospect now? First of all, I'll tell you what you can do after learning Python.

First, artificial intelligence.

Python is the golden language of artificial intelligence. It naturally chooses artificial intelligence as the employment direction, with good employment prospects and generally high wages. In Lagou.com, the starting salary of artificial intelligence engineers is generally 20K-35K. Of course, if they are junior engineers, the starting salary has already exceeded12,500 yuan/month.

Second, big data.

We are currently in the era of big data. Python is more efficient than Java in big data. Although big data is difficult to learn, Python can better connect with big data. The salary of big data with Python is at least 20 K, and big data continues to be hot. In the future, the salary of big data engineers will gradually increase.

picture

Everyone will certainly encounter many problems when learning python, as well as the pursuit of new technologies. Here we recommend our python learning button QUN: 784-758-2 14, which is the gathering place for Python learners! ! At the same time, I am a senior python development engineer, from basic python scripts to web development, reptiles, django, data mining and so on. , sort out all the data from zero foundation to actual project. Give it to every python friend! Share some learning methods and small details that need to be paid attention to every day * *

Third, the network crawler engineer

As a sharp weapon of data collection, web crawler is very useful as a source of data in the era of big data. Using Python can improve the accuracy and speed of data capture faster, which is the well-being of data analysts. Through the web crawler, BOSS no longer has to worry that you have no data. The salary of being a reptile engineer starts at 20 K. Of course, because of big data, the salary will rise all the way.

Fourth, Python Web full-stack engineer.

Full-stack engineer refers to a person who has a variety of skills and can use these skills to complete products independently. Also known as full-end engineer (with both front-end and back-end capabilities), English full-stack developer. Full-stack engineers are all talents in any language, and the salary of Python Web full-stack engineers is basically 20K higher, so if you have enough ability, Python Web full-stack engineers are the first choice.

Verb (abbreviation for verb) Python automates operation and maintenance.

Operation and maintenance workers have a great demand for Python, so act quickly. Learning Python automatic operation and maintenance can also get the salary of 10k- 15k, which is very good.

Six, Python automated testing

Python is a very efficient language. As long as it is related to automation, it can play a huge advantage. At present, most workers who do automated testing need to learn Python to help improve testing efficiency. Testing with Python can also be said to be a necessary tool for testers. The starting salary of Python automated testing is generally around 15K, so test partners also need to learn Python!

Seven, 3D game development

Python has a good 3D rendering library and game development framework, and there are many practical Python-developed games, such as Disney Animation City and Blade of Darkness. Commonly used PyGame, PyKyra and so on. And a PyWeek game. Python is also a good choice for students who want to enter the game industry.

Eight, business technology architecture evaluation and optimization

The quality of the code itself is enough to affect the access efficiency, which is difficult to improve through the optimization of clusters and servers the day after tomorrow. The ability to develop can make it possible to evaluate whether the technical architecture is reasonable and where it can be adjusted. The ability to develop, design and optimize architecture is a necessary ability to be a good architect.

Yes, there is no doubt that python+numpy+matplotlib is enough to replace it, not to mention other packages.

I think at least in China, it is the general trend that matlab is gradually replaced by python. The reason for this is the following：

1)matlab is a commercial tool launched by mathworks Company in the United States, with the emphasis on "American companies". Since ZTE and Huawei used to be harmed by the United States, who dares to put their lives on American companies? God knows when they will be banned for no reason. Python is certainly available, gradually replacing it.

2)matlab is commercial software, and lisense is expensive. Moreover, mathworks has left many backdoors in the software to report user information. Regular companies feel heartache when using genuine software, and fear when using pirated software. Since there is free python, why not? Even if the function is weak, we should believe that the power of the community is infinite and can make up for it soon.

Matlab and python are not on the same level. Matlab is a product oriented to the algorithm itself and the simulation itself. If we have to talk about running efficiency, it depends on who wrote the program. The reason why matlab charges lies in its runtime update. For example, updating the 5NR library in time is not impossible if it is written in python, but it is difficult to guarantee timeliness, integrity and running efficiency. After all, behind matlab is a strong team of scientists responsible for the algorithm, and a strong team of engineers to complete the implementation, and finally give users a simple and easy-to-use function. Users do algorithm simulation and realize their own algorithms. Everyone has done what he is best at.

I don't think so. Python can't do many professional simulations.

Matrix thinking, matrix visualization and concise grammar are all lacking in python.

Matlab focuses more on algorithm research and simulation. Python is a hodgepodge. Personally, I think Matlab is more suitable for debugging algorithm details. There is also that Simulink can't be completely replaced in many fields for the time being.