bda (version 3.1.3-2)

bde: Density Estimation for Binned Data

Description

Returns x and y coordinates of the density estimate of the probability density based on binned data.

Usage

bde(f, breaks, dist="ewd",gridsize=512L,iter=100,...)

Arguments

f

vector of frequencies (counts) of different bins. Missing values are not allowed.

breaks

vector of breaking points.

iter

iteration for model fitting.

gridsize

the number of equally spaced points at which to estimate the density.

dist

Distribution family to be fit.

...

controls

Value

• a list containing the following components:
• x
  vector of sorted x values at which the estimate was computed.
• y
  vector of density estimates at the corresponding x.
• mean
  an estimate of the mean.
• sigma
  an estimate of the standard deviation.

encoding

UTF-8

Details

A specific family of distribution is fitted to the a set of non-negative data that have binned. Missing values are not allowed. Families of distributions supported include: 'ewd' -- exponentiated Weibull distribution; 'weibull' -- Weibull distribution; 'gb' -- generalized beta distribution, 'dagum'. Other ditributions being developed include 'gld', 'normal', 'gamma', 'beta'.

References

Wang, B. (2013?). A manuscript is in preparation.

Examples

## Mobile 2010 Household income
## (http://www.clrsearch.com/Mobile-Demographics/AL/Household-Income)
##  reported mean $61291, median $42574.  

 breaks <- c(0,15000,25000,35000,50000,75000,100000,125000,150000,200000,Inf)
 x <- c(15783,9748,9150,11512,13609,9422,4861,2565,1843,2495)

## par(mfrow=c(2,2))
 (out1 <- bde(x,breaks, dist='gb'))
## (out2 <- bde(x,breaks, dist='weibull'))
## (out3 <- bde(x,breaks, dist='ewd'))
## (out4 <- bde(x,breaks, dist='dagum'))
 hist(out1)
 lines(out1, col=2, lwd=2)
## hist(out2)
## lines(out2, col=2, lwd=2)
## hist(out3)
## lines(out3, col=2, lwd=2)
## hist(out4)
## lines(out4, col=2, lwd=2)