R: Weighted Histogram

whist {heR.Misc}

R Documentation

Weighted Histogram

Description

Function adapted from VR's truehist function in MASS library to calculate a weighted histogram

Usage

whist(data, w = NULL, nbins = NULL, h, x0 = -h/1000, plot=TRUE, 
      axes=TRUE, breaks, prob = TRUE, xlim = range(breaks), ymax = max(est),
      col = 1, xlab = deparse(substitute(data)), bty = "n", main = NULL,
      verbose=T,  ...)

Arguments

`data`	raw data from which to calculate the histogram
`w`	vector of weights with same length as the data vector, if `NULL`, then the usual histogram is produced
`nbins`	number of bins
`h`	bin width, takes precedence over `nbins`
`x0`	shift for the bins
`plot`	logical, whether to produce the plot or not
`axes`	logical, whether to include the axes or not
`breaks`	actual break points, takes precendence over `nbins`
`prob`	logical, whether to plot probabilities or frequencies
`xlim`	two-element vector giving the x axis limits
`ymax`	the upper limit for the y axis
`col`	the color to use for histogram rectangles
`xlab`	the label for the x axis
`bty`	the box type for the histogram plot
`main`	the title of the plot
`verbose`	logical, if `TRUE`, then output the number of bins and data size after `NA` removal
`...`	additional graphical parameters

Details

Deletes all records that have missing values in either data or w.

Due to an issue with factor, this won't work if there are zero counts in any given class interval.

This function is based on truehist by Brian Ripley in the MASS library.

Value

A list of histogram breaks, probabilities or weighted probabilities and sum of weights if weights are specified, normalized probabilities or weighted normalized probabilities, and the raw counts

Author(s)

Neil E. Klepeis
neil AT exposurescience DOT org
http://klepeis.net