"Big data" and visualizations make everything look simple, but #COVID-19, like everything else in the real world, is messy. OP argues that the data is noisy, contradictory, and incomplete ... and that it will be years before we can reliably draw any conclusions from it.
After the country made vaccines illegal vaccine related death declined. In fact if you compare all countries where vaccines are illegal you will find vaccine related death is almost non existent. But in other countries where vaccines are legal vaccine related death is in the hundreds every year. I think this clearly shows why vaccines kill and should be made illegal.
This is what people who are anti-gun sound like. Do they even realize how absurd and how much lack of understanding they have when it comes to interpreting data and statistics?
I might try my hand at it after this semester to (1) practice with #Python, (2) use #Beeware [1] some more, and (3) build a data set of my posts that I can run #textAnalysis and #statistics tools on for fun and to get an idea of what content I can easily add to my #wiki.
I noticed an interesting statistical pattern thing today that my brain seemed to have done while I was doing my job. I'll try to write up a description of it shortly. π My mindfulness meets my growing math-y intuition! #personal#math#statistics
If anyone with some #statistics or #mathematical inclination is looking for a good way to get into computery stuff, I heartily recommend Data Analysis With Open Source Tools. It seems like a good way to apply modern stats to large data sets.
I wish I could recommend it to computer people wanting to learn more mathematics, but I think the mathematics in the book might be scary, since it's not often explained from first principles. It seems better for the other direction.
@sivy 1M records isn't small. Traditional stats is made for data sets where 50 observations is a generous size.
So, since you're saying "records", I am guessing you do database stuff so you know how to program. A popular book for your presumed ilk is Think Bayes. I haven't read it, but heard good things.
@Maenad Last semester for Halloween @batjan's #statistics prof attached two tails to her butt and was handing out tea bags at the beginning of the class. Once the undergrads settled down for the lecture to begin, she explained:
When getting a spot on campus to chat with @zacts I happened to sit next to a group of guys that ended up debating #stats about basketball, comparing stats about present day players versus classic era (Michael Jordan, Magic Johnson). Knowing little basketball data but some statistics, it was fun to infer what basketball history they were likely talking about. At one point I interjected to defend a guy whose point wasn't being heard, and my contribution was appreciated. π #statistics#college
Future project: predicting #Liberapay stats, week by week, determining whether there's seasonality (not necessarily dealing with seasons but whether, say, donations tend to be given in the first week of the month while withdrawals are done in last week of month). #stats#statistics#projects
@baldur wat they're measuring book sales by gross revenue instead of numbers of books sold. What if it's just the case that book prices declined? If that is the case and they've declined very significantly, numbers of #books sold might actually be increasing. #statistics