calculate mean from histogram

# Code to illustrate how to calculate mean of a histogram (possibly
# with uneven class-widths).

# Example data:
# As input we use a frequency distribution of age groups, stored in two variables:
#   nr = the counts per age group
#   agegrp = the labels for each age group

nr <- c(8, 11, 26, 20, 13, 5)
agegrp <- c('<1','1-2','3-5','6-11','12-17','18+')

# Based on the variable agegrp, the mean age of each age-group is
# determined. The result is stored in the variable midgrp.
# The mean age is just the middle value of the range - here done by
# 'hand-work' (no programming).
# For the highest interval, the assumption was made that it ranges up to 20.
midgrp <- c(0.5, 1.5, 4, 8.5, 14.5, 19)

# Now the estimated mean age for this frequency distribution
# is calculated by multiplying the mean per interval with the reltive frequency
# per interval and summing these values.
relfreq <- nr/sum(nr)
( mean <- sum(midgrp * relfreq) )