User:Mr.Z-man/Popular pages FAQ

From Wikipedia, the free encyclopedia
Jump to: navigation, search

This page summarizes the most commonly asked questions about the popular pages system. If you have any further questions, please ask.

Questions[edit]

1. How do I set one up for a project I'm involved with?
You can make a request on Tool Labs, here. Please don't request reports for projects you aren't actually affiliated with, Tool Labs is a shared resource, and if a project isn't actually interested, its just a waste of resources.
2. Is my project already set up to make a list?
You can see all of the projects that are signed up here (coming soon)
3. If my project is on the list, why hasn't the page been created yet?
The bot only creates and updates each page once per month. Note that the bot won't overwrite an existing page when it creates it, so do not put a placeholder on the page, or it may not be created.
4. How will I know when the page is created?
The bot will post a message to the project's talk page after creation.
5. Can I edit the list page after the bot creates it?
Yes, with some limitations. Any edits to the list itself will be overwritten each month, so edits should only be made to the header area. Do not change the name of the section header for the list section or it may break updates.
6. Why is the assessment information for some articles wrong?
The assessment data is recorded for each page once per month, at a time that can't be predicted. So if the assessment changed during the month, the bot may still list the old assessment.
7 Why is the data different from other sources such as stats.grok.se?
Most other programs simply compile the raw data from the WMF's text files into some more easily-queryable format. This tool goes a step further and aggregates data from redirects under their target pages. The raw data itself can often be difficult to parse, so there may also be differences in how it is "interpreted" regarding things like encoding.
8. Is the data completely accurate?
Not really. Occasionally, some of the raw data the bot uses is simply missing – this is typically only an hour or two of data per month, but rarely it can be almost an entire day's worth. In these cases it's filled in with other data from the month to approximate the correct results. There have been issues in the past with not all views being counted. And the process used to collect the raw data does not distinguish between human users and bots such as search engine crawlers.
9 Is data available for other languages?
Not at this time. If you're interested in helping with this, see "How can I help?" below.

For developers[edit]

10. Is the data available in some form for me to do my own analysis?
Yes. The data used by the bot to generate the lists is in a database on the same server as enwiki on Tool Labs. Contact Mr.Z-man for details. From January 2014 onward, it includes data for basically all articles on the site; earlier months include only a subset of pages. The raw data files are here.
11. How can I help?
If you're interested in improving or maintaining the bot, contact Mr.Z-man. The code is mostly Python (for the backend and the bot itself) and PHP/Javascript (for the web tools). The program that actually parses the raw data is C and the data is stored in MySQL tables. There has been some interest in expanding coverage to other languages, so experience in internationalization and localization would be especially helpful.

Other stuff[edit]