"College students who know python, how much does it make to find a job?"
"What the hell is python? I relied on him to get 5 offers"
"Is Excel still used for data analysis? Learning python is 10 times more efficient!"
......
You must have seen such a title. The official account of "python data analysis" or articles are everywhere. Of course, if you click on an article similar to the above title, it must be a python course registration at the end...
Under the brainwashing of overwhelming advertising, many people who want to learn data analysis or have been engaged in data analysis have this question: Is python really good? Do I need python to learn data analysis? Why can it be so hot?
First of all, it is undeniable that python is indeed a very good programming language with a wide range of applications, simple syntax, clear algebraic logic, and a large number of third-party libraries.
It has very powerful functions in the field of data analysis, data crawling, cleaning, visual analysis, mining...python is omnipotent. The development efficiency is high, the operation speed is fast, and the entry is simple. It is said that the information courses for elementary school students in some areas have even added Python. It seems that learning Python is the general trend and it is unstoppable.
But do you have to learn python for data analysis?
The answer is obviously no.
Whether it’s python, R, Excel, or spss, these are tools for data analysis. For data analysis, I have always emphasized that the core is business. Through business analysis logic, it is mapped to the processing logic of data analysis, and data analysis tools help us Means to achieve results.
If you compare the result of data analysis to a destination you are going to, then python is just a means of transportation that can reach this destination. In other words, you can do it with another tool, so between python and data analysis , And there is no inseparable relationship.
Since it is about choosing tools, it must be the best tool to achieve the goal the fastest. Is python the best choice for data analysis tools?
Not certainly. Different roads are suitable for different means of transportation. Similarly, for different types of data analysis tasks, suitable data analysis tools are also different.
In actual work, the job level of data analysis is different, and the job functions are also very different. In different companies, they are also called data analyst positions. Perhaps one is to obtain business data and provide basic data support. The other involves data modeling and mining. The tool skills that these two people need to master are definitely quite different.
Here I put the general classification of data analysis into two categories: business and technology :
Business analysts focus on business analysis, usually affiliated with a business department, or have a separate data analysis department. The most important task is to do special analysis of specific businesses, and do some business plans and plans through data analysis. Most of the daily work is to sort out reports, do some exploratory business analysis, and solve business problems.
Technical analysts are generally in the IT department and data center. According to different work links, they are divided into database engineers, ETL engineers, crawler engineers, algorithm engineers and other roles. The main tasks generally include data warehouse construction, special analysis, modeling analysis, data mining forecasting, etc.
After talking about the work content of data analysts, let’s look at several types of data analysis tools currently popular in the market: Excel, python/R, BI tools
Let me talk about Excel, which everyone is familiar with. Excel 's position in the field of data analysis is unshakable. Especially for beginners, most people have some contact with Excel before entering work, so on this basis, do data analysis and learn Excel. It is the most suitable. From simple table making, pivot table, writing formula, to VBA language, it can basically meet the analysis needs of 80% of business personnel. And the cost of learning Excel is much lower than learning a programming language.
(I have seen many organizations start to use workplace anxiety to fool ordinary workplace people to learn python. To put it conscientiously, they spend thousands or even more than 10,000 yuan to register for classes to learn python. Without basics for a few months, they can only learn python It’s not used at all in your own work, unless you seriously consider changing careers. Of course, changing careers cannot be as simple as the training organization said)
Back to the topic, let's talk about BI tools. The purpose of BI is to shorten the time from business data to business decision-making and improve decision-making efficiency. Therefore, its product design concept is developed around the process of improving data analysis.
Compared with Excel, BI tools are more simplified in the analysis process. Take the FineBI I have used as an example. From data linking, data processing, to visual chart analysis, many functions are encapsulated, and you can quickly click and drag. Complete an analysis. Such a visual operation interface greatly reduces the learning threshold of BI and is more suitable for business analysts in the enterprise.
Data processing
Visual analysis
In addition, in the face of large data analysis, BI tools can also make up for the shortcomings of Excel. Another attractive point is the visualization effect of BI tools. To make dynamic charts or advanced visualization chart effects in Excel, you need to go through Many complex steps are implemented in a programming language and require line-by-line code adjustments to get the desired effect. However, in the BI tool, simply drag and drop settings can create stunning visualization charts.
Visual cockpit made by FineBI
However, because BI tools are not open source, they have limitations in functionality. If a product is not designed for a certain function, there may be no way to complete the analysis. At this time programming languages such as python or R are more flexible. As long as the code is well written, there is basically nothing that cannot be achieved.
And in the field of data mining and machine learning, BI and Excel are even more powerless. At this stage, the python language is an excellent choice. It is more comfortable with complex data processing, modeling, and prediction.
Finally, to sum up, the choice of tools should be based on their own needs, not which one to learn from, only the best for you is the best.
For basic business analysis such as finance, personnel, and operations, excel is completely enough. If you want to improve efficiency and pursue visualization, BI tools are also a good choice. There is no need to spend a lot of energy to join in the excitement of Python. , Of course, if you are interested in programming, then I did not say.
Finally, I will answer the last question, why is python so popular?
Of course, it’s because it makes money. In the past, when the Internet emerged, various java and C++ trainings were hot, and training institutions made a lot of money. Now the era of big data has come, and the concepts of data analysis, data mining, and artificial intelligence have come again. It's on fire, and a new field of leeks appears in front of you. Can training institutions still keep so much money without making money?
Just take the salary of a TOP data analyst in the industry and give you a pie of 2W monthly income. In fact, you have a high probability of entering a job of 6-8K, and you have learned that little skill in the training course and face the whole network being cut. The gods of all walks of life of Leek have turned to PK with you, and in the end there are very few who can find their favorite jobs.
What’s worse is that some innocent little whites have done a good job in their original positions. Seeing the marketing copy, they have to switch on a whim. They spent a lot of money and energy to apply for classes to learn python. In the end, they couldn’t change jobs, but they didn’t. What's the improvement in my position, the dream of making big money is broken.
Finally, I want to emphasize that most data analysts are essentially business support positions. The core is business understanding and data sensitivity. As written in the following picture, those who tell you to learn python can get started with data analysis. Python can do a good job in data analysis, 90% of which are for making money. In addition, those who want to switch to data analysis should also consider carefully. This business is not as good as you think.
