Recently been listening to Jay Chou's songs (small series may be grown up, there may be spring is coming ....), then went to get an analysis of the data Python Jay songs.
But people familiar should know, Python crawling simple, but it is very difficult to analyze, SQL statements, Pandas and Matplotlib battle together, most people will not, learn it also takes a lot of time.
So I thought of a new way, Python crawling + BI data visualization analysis, the former fast speed, the latter step straightforward drag and drop, both with simply invincible!
First, what data analysis?
Jay ah! My youth! Dimensional analysis can be many.
I want to see every song he were listening for what season, what appeared most of the lyrics word again, and the composer and lyricist who are a little more romance for which song to listen to, what to listen to a song for love .. ..Too much!
Second, the process of crawling
First we enter into what we need to crawl address: Netease cloud music homepage, our aim is to grab all of Jay's songs, lyrics, and comments, then we enter in the search at Jay.
Get this picture, we find that there is only a maximum of 50 songs (many songs analysis Netease cloud will only select TOP50), all we want is, so this URL does not meet the requirements, we continue to look for other URL addresses.
Here I spent a lot of time, and finally found an indirect method, first grab all of Jay's album information, and then go looking for all songs by album information.
We look at what header inside, there's a string we do not control, as it has been in our url inside, we just look at the request headers thing is that we send to the server.
Code section:
Specific results are as follows:
Using the same method, we can analyze the information of the song, Lively information specific code at the end of the text, the final amount of data of about 30W ~
Third, the visual analysis data BI
In recent years, BI came out of the show, have emphasized the agile, self-service, a change in the traditional BI tools SAP BO, IBM cognos home monopolize the market situation, which is a good thing, it also proves the business intelligence field in the increased value .
How BI product selection?
Price: within the range of allowable expenses, so that the highest price.
The maturity of the product: high maturity and practical needs of most enterprises are located, there is a general maturity and proportional to the number of products and customers time, because the market has been inspected.
Rapid deployment of the product: It is important for fast on-line.
Development function: whether the friendly enough, if enough agility, business people can be used directly.
For the above, together with their own research and found Tableau and FineBI are two leading BI tool industry.
Tableau Visual Insight analysis is praise, but also with audio version, but because it is a foreign product, mostly dealers, service can not be guaranteed, and the operation mode is not suitable for use white; the same time, companies with multiple concurrent expensive.
FineBI domestic brands, but also one of the few able to occupy the world's leading data tool, comes with ETL, visualization is very cool, the main business personnel self-analysis, Enterprise Edition price is very conscience, free for personal use.
Fourth, Jay song data visualization
FineBI not only beautiful visual analysis, and its operation is very simple.
Some would say that Excel is not bad, I suggest you wake up. Excel data visualization do not work, the amount of data too large stuck, VBA programming is also required, the type of chart is also particularly special little .....
1. What sad song
Jay nunchaku Huobian year against the world, but careful analysis will find that, in fact, his songs are the main sad love songs, he analyzed where the emotions are more depressed a few songs, recommend lovelorn people do not listen to (funny face).
2. What are cheerful song
Of course, there is depression, there is delightful. For example, Tao Heung, is a very cheerful song, remember the home is the only castle, as the river continues to run .... Tao Heung small series called iron.
3, heat ranking
Because the songs are older, there will be no increase in the heat, but still a classic! "Say good cry," Jay's new song, the evening when the line led directly to the collapse of QQ music, you can see its influence.
4, season rankings
Jay will look at what season release, without analysis also know that in the fall, love songs are more likely to fall in the right mood (I'm guessing).
5, lyricist accounting
Jay song lyrics mention, we first thought is certainly Fang, but then take FineBI analysis found that Chou himself is a gifted lyricist.
V. Summary
The key is to use business management and BI products. Depth analysis of data related to the higher needs of the business thinking of allocation of resources, which is why FineBI flagship self-analysis, so business people to discover the value of data.
These are the first data analysis FineBI of combat, but also to help you better understand this product, but it is just above the tip of the iceberg, to welcome you to personally experience this enterprise-class data analysis tools.
