NTplot(mean0, ...)
## S3 method for class 'default':
NTplot(mean0=0, ..., shiny=FALSE,
distribution.name = c("normal","z","t","binomial"))
## S3 method for class 'htest':
NTplot(mean0, ..., shiny=FALSE, NTmethod="htest")
## S3 method for class 'power.htest':
NTplot(mean0, ..., shiny=FALSE, xbar=NA, ## these input values are used
mean1, n, df, sd, distribution.name, sub, ## these input values ignored
alpha.left, alpha.right, number.vars) ## these input values ignored
## NTplot(NTplot(htest.object), n=20) ## allows override of arguments
## S3 method for class 'NormalAndTplot':
NTplot(mean0, ..., shiny=FALSE)
mean0
is either missing or a numeric
argument for the mean under the null hypothesis.
For the htest
method, mean0
is an htest
object
from the
NormalAndTplot
.NormalAndTplot
.TRUE
, a shiny
app is started to provide an interactive graphics device in a
web-browser. If
FALSE
, a plot is drawn on the current graphics device.TRUE
for "htest"
objects.power.htest
method.htest
and
power.htest
methods.
Otherwise passed on to the next method.shiny=TRUE
. It is
normally calculated by the methods. NTmethod
tells
shiny
how to use or ignore the df
and n
sliders. See the extended discussion in"trellis"
object. The object can be plotted or fed back into the
NTplot
function with argument shiny=TRUE
to allow
interactive graphical investigation of the hypothesis test or confidence
interval.
The attributes of the object\
NTobj <- NTplot()
\
attr(NTobj, "scales")
and attr(NTobj, "prob")
make the
data values and probability values accessible for further R
computations.
The "call"
attribute
cat(attr(NT.object, "call"), "")
displays a statement that can be copied back into R to reproduce the graph.
The cat()
is needed to unescape embedded quotes.
The "call.list"
attribute
attr(NT.object, "call.list")
is a list that can be used with do.call
to reproduce the graph.
do.call(NTplot, attr(NT.object, "call.list"))
. This is usually
not needed by the user because the simpler statement
NTplot(NT.object)
does it for you.
NormalAndTplot
, print.NormalAndTplot
.normal.and.t.dist
built on base graphics that is used in many
displays in the book by Erich Neuwirth and me: R through Excel, Springer
(2009).
NT.object <- NTplot(mean0=0, mean1=1)
NT.object
attr(NT.object, "scales")
attr(NT.object, "prob")
cat(attr(NT.object, "call"), "
NTplot(t.test(x1, x2)) NTplot(power.t.test(power = .90, delta = 1, alternative = "one.sided"))
## 22 distinct calls are shown in demo(NTplot, ask=FALSE)
## these are interactive and do not work in static checking of the code NTplot(mean0=0, mean1=1, shiny=TRUE) NTplot(t.test(x1, x2), shiny=TRUE, mean1=1) NTplot(power.t.test(power = .90, delta = 1, alternative = "one.sided"), shiny=TRUE) NTplot(NT.object, shiny=TRUE)
t.test
, power.t.test
, and z.test
are accepted and displayed.NTplot
is built on xyplot
.
Most of the arguments detailed in xyplot
documentation work to
control the appearance of the plot.
The shiny app (called when the argument shiny=TRUE
)
provides animated sliders for the means, standard
deviation, xlimits, significance levels, df, and n. The df and n are
rounded to integers for the sliders (relevant for htest
and
power.htest
objects).
When you have a graph on the shiny window that you wish to keep, click on the "Display Options" tab, and then on the "Display Call" radio button. The main shiny window will show an R command which will reproduce the current plot. Pick it up with the mouse and drop it into an R console window.
To get out of the shiny window and return to an interactive R console,
move the cursor back to the console window and interrupt the shiny call, usually
by entering Ctrl-C
or ESC
.