youtube_subscriptions

YouTubers' weekly subscriptions and video views since the Covid19 pandemic

Setup

Steps for collecting url and name of the Youtuber of interest:

• Find the account name of the Youtuber as shown in their channel
• Search for this Youtuber in https://socialblade.com/youtube/
• In the homepage, click on "See full monthly statistics" to go to the detailed statistics page.
• Now the url should end with "/monthly". Copy this url and go to https://web.archive.org/.
• In the "save page now" section, enter the url you just copied and save the page.
• After the page is saved, you will see on the screen "visit page: [url]". Copy that link and put it into the 'url' column of the google sheet.
• Alternatively, you can add an chrome extension of the wayback machine (https://chrome.google.com/webstore/detail/wayback-machine/fpnmgdkabkmnadcjpehmlllkndpkmiak?hl=en-US). Then you can click the icon in the tool bar while you are on the socialblade monlty statistics page, and then click save page now, and click the "copy the archived link" icon in the bottom right corner to copy the url we need to the clipboard.
• The 'name' column on the google sheet is the account name as shown in the url, right before '/monthly'. NOT the youtube channel's name.
  • E.g. url: https://web.archive.org/web/20220521004032/https://socialblade.com/youtube/user/pamelarf1/monthly, name: pamelarf1.

Web scraping

Under code, there is a scraper_batch.py file which was used to batch scraping the websites after we collected the urls and names and compile them into the library.csv file under data. If you want to try scraping a single website, you can use the scraper_revised.py file. It will ask you to input the url and name of a single website you want to scrape. To avoid any conflicts, please do not push any scraped data to this repo until we finalized the list of urls.

Data cleaning and analysis

• The R code used for data cleaning and analysis could be found under code -> data_cleaning_analysis_R.Rmd, and the Python code in code -> data_analysis_python.ipynb.
• The raw scraped data organized in different folders based on YouTube channel categories could be found under data. Also, we included the cleaned data in the same "data" folder.

Report

A PDF version of the final report can be found under report.