Ten years of data analysis experience, summed up the best use of these three types of tools

Time:2019-12-9

When it comes to data analysis tools, I believe you are familiar with them, but many people have a doubt?

With so many data analysis tools, what’s the difference between them? Which is better? Which is stronger? Which should I learn?

Although this question is a bit conventional, but it is very important, I have been trying to pursue the answer to this ultimate question. If you search for relevant information on the Internet, it’s hard to see a fair point of view. Because the reviewers who evaluate a tool are likely to stand in different angles, with some personal feelings.
Ten years of data analysis experience, summed up the best use of these three types of tools
Ten years of data analysis experience, summed up the best use of these three types of analysis tools
Today, let’s put aside these personal colors and try to objectively discuss with you my personal views on several data analysis tools on the market for your reference.

I chose three types of tools:

Excel
BI tools
Programming languages such as R and python
Let me introduce:

Excel
If Bi tool is a fighter, R language and python are bombers, then excel is the aircraft carrier in data analysis, with powerful functions such as table making, PivotTable, VBA, etc. the system of Excel is so huge that no analysis tool can surpass it to ensure that people can analyze according to their needs.
Ten years of data analysis experience, summed up the best use of these three types of tools
Of course, some people think that they are very proficient in computer programming language and disdain to use Excel as a tool, because Excel can’t handle big data. But think about it differently. Do we use more data in our daily life than big data? In my opinion, excel is a versatile player. It is the best way to solve small data. Plus plug-ins, it can also process millions of data.

To sum up, based on the powerful functions of Excel and its user scale, my opinion is that it is a necessary tool. If you want to learn data analysis excel, it is absolutely the first choice, and it is a must!

BI tools
Bi is also called business intelligence, which is born for data analysis. Its starting point is very high. Its goal is to shorten the time from business data to business decision-making, and how to use data to influence decision-making.

But we don’t think Excel’s product goal is like this. Excel can do many things. You can draw a curriculum table, make a questionnaire, use it as a calculator to calculate, and even use it to draw, write a little game with VBA. These are not data analysis functions.

Ten years of data analysis experience, summed up the best use of these three types of analysis tools
But the technology industry has a specialty, Bi is specialized in data analysis.

Take the common Bi tools such as powerbi, finebi and tableau on the market. You will find that they are designed in full accordance with the process of data analysis, first data processing, sorting and cleaning, then data modeling, finally data visualization, showing charts, telling stories with graphs, and exploring problems affecting decision-making.

These are the only ways for data analysis. At the same time, there are some pain points for practitioners in this process:

For example, the repetitive and low value-added work of cleaning data can be simplified with Bi tools;
Do data perspective analysis, because of the large amount of data, the traditional excel tools are very difficult, card, crash;
It may take a lot of time to edit the chart with Excel, including setting the color and font;
These pain points are the places where Bi tools can bring us change and value-added.

Let’s talk about the comparison between powerbi, finebi, tableau and other Bi tools:

1、Tableau:

In fact, the core essence of tableau is Excel’s PivotTable and PivotChart. It can be said that it has been acutely aware of this PivotTable feature of Excel, cut into the Bi market earlier, and carried forward this core value.
Ten years of data analysis experience, summed up the best use of these three types of tools
Ten years of data analysis experience, summed up the best use of these three types of analysis tools
In terms of development history and current market feedback, tablueau is better at visualization. This advantage I think is not how cool the chart is, but its design, color and operation interface give people a simple and fresh feeling. This is exactly what tableau publicized. It has invested a lot of academic energy in studying what kind of charts people like and how to bring the ultimate experience to users in terms of operation and vision.

In addition, tableau is becoming more and more perfect, such as adding data cleaning function and more intelligent analysis functions. This is also tableau’s predictable product development advantages.

2、Power BI

Power Bi wins the data analysis function of Microsoft’s business model and products:
Ten years of data analysis experience, summed up the best use of these three types of tools
Before powerbi, excel plug-in was used as a product. Limited by the aircraft carrier of Excel itself, the development situation was not ideal, so powerbi was separated from excel plug-in, independent into a school and completely transformed. But as a latecomer, there are iterations and updates every month, and the catch-up speed is very fast.

Ten years of data analysis experience, summed up the best use of these three types of analysis tools
The business model of powerbi is software free, so you don’t have to worry about piracy and version cracking, because the genuine version is free, which is very attractive compared with the price of tableau of thousands of yuan; on the other hand, the data analysis function is PowerPivot, DAX language, which allows me to write formulas in a way similar to excel, to achieve many very complex advanced Analysis.

3、Fine BI

As for fine Bi, its unique feature is that self-service Bi is more suitable for enterprise users.
Ten years of data analysis experience, summed up the best use of these three types of tools
For example, the business personnel will have a demand for data retrieval. The data here is not correct, and the report format there is not correct. The efficiency is very low. As some enterprises do not have data analysts, the self-service of finebi can realize data retrieval and analysis within the authority, and no longer allow business and it to argue with each other.

Traditional Bi methods may require ETL architects or data modelers, but self-service Bi requires little, basically can complete the liberation of labor and reduce costs as much as possible.

Another important point is that finebi implements data perspective analysis by dragging and dropping fields. It can generate charts with one click. The threshold for entry is relatively low. For novices of data analysis, it is better to learn than powerbi and tableau.

Ten years of data analysis experience, summed up the best use of these three types of analysis tools
R language and python
The third kind of tool, which is the most difficult to answer. Although the design of software such as Excel and Bi tools has tried their best to consider most of the application scenarios of data analysis, they are all customized in nature. If you do not design a function or develop a button for a function, you may not be able to complete your work.

For this point, programming language is different. Language is very powerful and flexible. You can write code to execute what you want. For example, R and python, as the necessary tools for data scientists, are definitely higher than excel and Bi tools from a professional level.

So what application scenarios can r and python do, while excel and Bi tools are difficult to implement?

1. Professional statistical analysis
Ten years of data analysis experience, summed up the best use of these three types of tools
Ten years of data analysis experience, summed up the best use of these three types of analysis tools
In terms of R language, it is best at statistical analysis, such as normal distribution, clustering by algorithm, regression analysis, etc. This kind of analysis is like treating data as an experiment, which can help us answer the questions:

For example, is the distribution of data normal, triangular or other types of distribution? How about dispersion? Is it within the statistical control we want to achieve? What is the magnitude of the effect of different parameters on the results? And hypothetical simulation analysis, if a certain parameter changes, how much impact will it bring?

2. Individual prediction and analysis
Ten years of data analysis experience, summed up the best use of these three types of tools
For example, we want to predict a consumer’s behavior, how long he will stay in our store, how much he will consume, or judge his personal credit situation through a person’s Taobao consumption record, and make a loan limit; or push different products according to your browsing record on the web page. This is also related to the current relatively popular concepts of machine learning and artificial intelligence.

Ten years of data analysis experience, summed up the best use of these three types of analysis tools
summary
The above comparison shows the differences of several softwares. I want to summarize that existence is reasonable. Excelbi programming language, these tools in the application of cross overlap, there are complementary places. For overlapping places, no matter what kind of tool, as long as you can use it to solve the problems you encounter, it is the best.

To choose which tool, you should first understand your own work and whether you will use the application scenarios I just mentioned. Or think about your career, whether it’s data science or business analysis.