crossing(x=seq(-1,1,.01), y=seq(-1,1,.01)) %>% mutate(sd = 0.005 + .005 * (x^2 + y^2 < 0.5)) %>% mutate(dy = rnorm(n(), 0, sd)) %>% ggplot() + geom_line(aes(x=x, y=y+dy, group=y)) + theme_void()
Mapping noise intensity to geometric properties can generate cool visual effects. See also here.
library(tidyverse)
crossing(x=seq(-1,1,.01), y=seq(-1,1,.01)) %>% mutate(sd = 0.005 + .005 * (x^2 + y^2 < 0.5)) %>% mutate(dy = rnorm(n(), 0, sd)) %>% ggplot() + geom_line(aes(x=x, y=y+dy, group=y)) + theme_void()