epiDisplay (version 3.5.0.2)

Follow-up Plot: Longitudinal followup plot

Description

Plot longitudinal values of individuals with or without stratification

Usage

followup.plot(id, time, outcome, by = NULL, n.of.lines = NULL, legend = TRUE, 
    legend.site = "topright", lty = "auto", line.col = "auto", 
    stress = NULL, stress.labels = FALSE, label.col = 1, stress.col = NULL, 
    stress.width = NULL, stress.type = NULL, lwd = 1, xlab, ylab, ...)

Arguments

id

idenfication variable of the same subject being followed up

time

time at each measurement

outcome

continuous outcome variable

by

stratification factor (if any)

n.of.lines

number of lines (or number of subjects in the data frame) randomly chosen for drawing

legend

whether a legend will be automatically included in the graph

legend.site

a single character string indicating location of the legend. See details of ?legend

lty

type of the "time" lines. See 'lty' in ?par

line.col

line colour(s) for non-stratified plot

stress

subset of ids to draw stressed lines

stress.labels

whether the stressed lines should be labelled

label.col

single integer indicating colour of the stressed line labels

stress.col

colour values used for the stressed line. Default value is '1' or black

stress.width

relative width of the stressed line

stress.type

line type code for the stressed line

lwd

line width

xlab

label for X axis

ylab

label for Y axis

...

other graphic parameters

Details

'followup.plot' plots outcome over time of the individual subjects.

If a stratification variable 'by' is specified, the levels of this variable will be used to color the lines.

'n.of.lines' is used to reduce the number of lines to allow the pattern to be seen more clearly.

'legend' is omitted if 'n.of.lines' is not NULL or the number of subjects exceeds 7 without stratification.

'line.col' works only for a non-stratified plot. It can be a single standard colour or "multicolor".

Values for 'stress.col', 'stress.width' and 'stress.type', if not NULL, should follow those for 'col', 'lwd' and 'lty', respectively

See Also

'plot','lines'

Examples

# NOT RUN {
.data <- Indometh
attach(.data)
followup.plot(Subject, time, conc)
followup.plot(Subject, time, conc, lty=1:6, line.col=rep("black",6))
detach(.data)

.data <- Sitka
attach(.data)
followup.plot(tree, Time, size)
followup.plot(tree, Time, size, line.col = "brown")
followup.plot(tree, Time, size, line.col = "multicolor")
followup.plot(tree, Time, size, n.of.lines=20, line.col = "multicolor")

# Breakdown of color by treatment group
followup.plot(tree, Time, size, by=treat)

# The number of lines reduced to 40
followup.plot(tree, Time, size, by=treat, n.of.lines=40)

# Stress some lines
length(table(tree)) # 79 trees followed up

# Identifying trees that sometimes became smaller
.data <- .data[order(.data$tree, .data$Time),]
detach(.data)
attach(.data)
next.tree <- c(tree[-1], NA)
next.size <- c(size[-1], NA)
next.size[tree != next.tree] <- NA
smaller.trees <- tree[next.size < size] 
followup.plot (tree, Time, size, line.col=5, stress=smaller.trees, 
  stress.col=2, stress.width=2, stress.type=2)
followup.plot (tree, Time, size, line.col=5, stress=smaller.trees, 
  stress.col=2, stress.width=2, stress.type=2, stress.labels=TRUE)
detach(.data)
rm(list=ls())  
  
# }