>>30489297It depends on the source of your data. Sometimes, for as stupid as it sounds, you just need to sit there and manually count it. To do that, you first figure out what values you want to pay attention to, so for your video example the three values you want to keep track of are:
1- Time, in this case date
2- Amount, in this case number of subscribers
3- Subject, in this case channel owners.
Then you need to find a way to measure your data. Since we're talking about a past action in the case of this video, they would probably use something like SocialBlade (I just checked and that video does use it). SocialBlad automatically logs popular channel's information every so often, so going back in time is relatively easy. You may need a paid account to go back really far, as I can't figure out how to. For something in real time, that's a much harder question to answer as that's more so based on your experimental set up.
However, the data that video uses is not as scary as you think. It's not filled with thousands upon thousands of data points per each second. They're most likely using a simple "lerp to" function to make it go from 60,000 to 60,940. You can think of it as a simple counting function. Something like [While I is in Range(0,939)-> Count= 60000 + i, print(Count)].
It's a bit hard to answer your question fully since it's really broad. How you get data is defined by your experiment. Your experiment defines what type of data you can collect. They go together. As a result, it's important to keep track of what other aspects of your experiment will interfer with your data. It may not always be obvious. I'm also not at my peak since it's like 3 am where I'm at, so I'm sorry if my answers aren't the clearest.
The Sana graph is really simple, but if you'd like to see it then here's the catbox link:
https://files.catbox.moe/ekhuzz.mp4Attached is a general set up for how the data is saved. Sadly, I don't have Sana's stuff saved anymore so most of it is useless to that video.
