Unconsciously, it has been more than a year since I wrote a small official account, and I still have a little understanding of the status of my official account.
This time, from a data point of view, the editor will take a look at the operation status of the editor over the past year.
Since the editor insists on originality and the number of articles is not too much, there is no crawler to crawl the data, but the data is manually entered.
Two tables of data, article information table, reprinting status table:
Through these two tables, perform exploratory description analysis, and then drill down analysis, the requirements are as follows:
Kinoshita learn Python
After more than a year of operation, a total of 64 articles have been published. Readers who have followed for a long time know that the editor publishes about two articles per month, so there are only so many articles in total:
What is the proportion of original, reprinted, and advertising in these articles? Readers who have followed for a long time know that the editor still insists on originality more:
In these articles, what is the wording situation of the title of the article? The editor made a word cloud in the shape of a file, and found that Python, crawlers, analysis, and data are more informative:
What is the number of users of the editor? The editor is just a small trumpet owner, does not load a lot, insists on originality, so the growth in the early stage is relatively slow, and after accumulating some quality articles, it grows faster:
The number of readings can reflect the overall distribution of the number of readings. To a certain extent, it reflects that the higher the number of readings, the more interesting the article is for everyone. The editor divides the number of readings into several intervals, mostly between 150-450 :
Not only the number of readings, the number of glances, the number of likes, and the amount of appreciation also reflect the reader's interest in the article to a certain extent. The editor checks the distribution of all articles and sees that each article is mostly 5, and the number of likes is 1- Between 20, appreciation is mostly distributed between 0-6 yuan:
The most popular article category in the editor is reptiles, because at first it is reptile articles, followed by others, including advertisements, tools, system environment, recommended books, self-summaries, article summaries, etc.:
The number of readings, take a look, like, admire, reprinted is the number of readings and take a look, the more you read, the number of take a look may also increase:
So what type of articles have a lot of reading? The editor has the highest number of crawlers. Then think about it, I always wrote crawlers in the early stage, and only started to write after data analysis. The number is not large, and the number of readings is relatively small:
The editor sometimes shares some group chats. Does sharing affect the number of readings? From the results, the number of shared readings will be relatively higher:
Several articles of the editor are of some quality, and have been reprinted by some bigwigs. The most one has been reprinted 9 times:
Let's take a look at the number of readings of the original text that was reprinted. Compared with the number of readings after reprinting by the big guys, the trumpet owner is always the trumpet owner:
The reprinted official accounts have these, and some of them are reprinted more than once, which seems to be quite good:
Articles with a high number of readings, the editor’s official account reading greater than 400 is considered high, after all, the trumpet owner, which type of article is the most? The most can reflect the reader's interest in this type of article to a certain extent:
So far, the editor has briefly analyzed the operation status, and I want to know more to try. The editor provides data sets and codes for your reference.
Source code acquisition
Follow the WeChat public account " Kinxia learn Python ", reply to " Kinxia " to get