R/fit.gaussian.R

In bayesball/LearnEDAfunctions: Functions for Learning Exploratory Data Analysis

fit.gaussian <- function(data, bins, g.mean, g.sd){
# bin data
h <- hist(data, breaks=bins, plot=FALSE)

# compute probabilities of bins
N <- length(bins)
probs <- pnorm(bins[2 : N], g.mean, g.sd) - 
  pnorm(bins[1 : (N - 1)], g.mean, g.sd)

# compute expected bins
expected <- probs * length(data)

# compute simple rootogram residuals
residual <- sqrt(h$counts) - sqrt(expected)

data.frame(counts = h$counts,
     probs = probs,
     expected = expected,
     residual = residual)
}