问题
I have some discrete data, which I have plotted in a histogram. I'd like to overlay a Poisson distribution to show the data is roughly Poisson distributed. Imagine the two plots from the code below merging into one plot, that is what I'd like to achieve.
# Read data
data <- read.csv("data.csv")
# Plot data
hist(data, prob=TRUE)
# Plot Poisson
c <- c(0:7)
plot(c, dpois(c, mean(data)), type="l")
I have tried the curve function:
curve(c, dpois(x=c, lambda=mean(data)), add=T)
But all I get is this:
The Poisson curve just seems to abruptly stop, but I would expect it to follow the shape of the histogram.
I'd like it to look like this (not necessarily with colours or multiple data sets):
回答1:
The code below does what you want.
set.seed(12111978)
vec <- rpois(50, 3)
hist(vec, prob=TRUE, ylim = c(0, .25)) # may need to tweak the y axis.
lines(0:max(vec), dpois(0:max(vec), mean(vec)), col = 'red')
来源：https://stackoverflow.com/questions/54112224/r-overlay-poisson-distribution-over-histogram-of-data