multi.hist {psych}

Multiple histograms with density and normal fits on one page

Description

Given a matrix or data.frame, produce histograms for each variable in a "matrix" form. Include normal fits and density distributions for each plot.

The number of rows and columns may be specified, or calculated. May be used for single variables.

Usage

multi.hist(x,nrow=NULL,ncol=NULL,density=TRUE,freq=FALSE,bcol="white",
      dcol=c("black","black"),dlty=c("dashed","dotted"),
      main=NULL,mar=c(2,1,1,1),breaks=21,global=TRUE,...)
histBy(x,var,group,data=NULL,density=TRUE,alpha=.5,breaks=21,col,xlab,
            main="Histograms by group",freq=FALSE,...) 

Arguments

x

matrix or data.frame

var

The variable in x to plot in histBy

group

The name of the variable in x to use as the grouping variable

data

Needs to be specified if using formula input to histBy

nrow

number of rows in the plot

ncol

number of columns in the plot

density

density=TRUE, show the normal fits and density distributions

freq

freq=FALSE shows probability densities and density distribution, freq=TRUE shows frequencies

bcol

Color for the bars

dcol

The color(s) for the normal and the density fits. Defaults to black.

dlty

The line type (lty) of the normal and density fits. (specify the optional graphic parameter lwd to change the line size)

main

title for each panel will be set to the column name unless specified

mar

Specify the lower, left, upper and right hand side margin in lines – set to be tighter than normal default of c(5,4,4,2) + .1

xlab

Label for the x variable

breaks

The number of breaks in histBy (see hist)

global

If TRUE, use the same x-axis for all plots

alpha

The degree of transparency of the overlapping bars in histBy

col

A vector of colors in histBy (defaults to the rainbow)

...

additional graphic parameters (e.g., col)

Details

This allows for quick summaries of multiple distributions. Particularly useful when examining the results of multiple-split halves that come from the reliability function.

By default, will try to make a square plot with equal number of rows and columns. However, the number of columns and rows may be specified for a particular plot.

Author(s)

William Revelle

See Also

bi.bars for drawing pairwise histograms and scatterHist for bivariate scatter and histograms. densityBy, violinBy and violin for density plots.

Examples

multi.hist(sat.act) 
multi.hist(sat.act,bcol="red")
multi.hist(sat.act,dcol="blue")  #make both lines blue
multi.hist(sat.act,dcol= c("blue","red"),dlty=c("dotted", "solid")) 
multi.hist(sat.act,freq=TRUE)   #show the frequency plot
multi.hist(sat.act,nrow=2)
histBy(sat.act,"SATQ","gender") #input by variable names
histBy(SATQ~ gender, data=sat.act) #formula input