Let’s say we have 2 points, (3,0) and (2,8) To find the linear line we would put both of them into equations, like this \(0 = a(3)+b\)
\(8 = a(2)+b\)
2024-02-15
Let’s say we have 2 points, (3,0) and (2,8) To find the linear line we would put both of them into equations, like this \(0 = a(3)+b\)
\(8 = a(2)+b\)
We then would solve those equations \(b=-3a\)
\(8=2a-3a\)
\(8=-a\)
\(a=-8\)
We then put the a value into one of the equations \(0=(-8)*(3)+b\)
\(b=24\)
So, our regression line is y=-8x+24.
Linear Regression is a similar concept, but the equation you get is not always accurate for all points. Instead, it finds the line that fits all of the points the closest. Let’s look at some examples.
This slide shows the year the movie was released compared to the rating it recieved for documentaries.
## `geom_smooth()` using formula = 'y ~ x'
This slide compares the percentile of 1 ratings compared to the percentile of 10 ratings for each movie
ggplot(movies, aes(x=r1, y=r10)) + geom_point(color = 'cornflowerblue') + geom_smooth(method = "lm", color = "violet")
## `geom_smooth()` using formula = 'y ~ x'
The graph on the previous slide got cut off because of the code, so this is that graph.
## `geom_smooth()` using formula = 'y ~ x'
This slide compares the length of short movies to the number of votes it recieves.