pathCEF {adaptTest} | R Documentation |
This function plots several conditional error functions of the same family such that each one runs through one of several given points.
pathCEF(typ = NA, fun = NA, dis = NA, p1 = 1:49/50, p2 = p1, x = 0:200/200, plt.pt = FALSE, plt.ptann = FALSE, xlab = NA, ylab = NA, ...)
typ |
type of test: |
fun |
a conditional error function |
dis |
a distortion method for a supplied conditional error function (see details): |
p1 |
a vector (at least of length 2) of p-values p1 of the test after the first stage |
p2 |
a vector (at least of length 2) of p-values p2 of the test after the second stage, defaults to |
x |
vector on which the conditional error functions are plotted (should be relatively dense in [0,1]) |
plt.pt |
logical determining whether the points that the conditional error functions are made to run through should be plotted or not (default: not) |
plt.ptann |
logical determining whether the points that the conditional error functions are made to run through should be annotated or not (default: not) |
xlab |
a label for the x axis (default: no label) |
ylab |
a label for the y axis (default: no label) |
... |
arguments to be passed on to the underlying |
It can be instructive to plot not only one conditional error function, but to visualize a whole family. This can easily be done with pathCEF
. The function is used in a similar way as plotCEF
, but p1
and p2
are now vectors (of the same length, at least of length 2). Conditional error functions are plotted that run through the specified elementwise points (p1,p2)
(which by default lie on the main diagonal).
Internally, pathCEF
uses plotCEF
to plot the individual conditional error functions; see this latter function for further details.
The function pathCEF
is invoked for its plotting effect; it returns no meaningful value.
Provide either typ
or fun
, not both! If fun
is provided, then also specify dis
.
Unlike plotCEF
, it is not possible with pathCEF
to specify the conditional error functions by the parameter alpha2 or the parameter c.
plt.ptann
is not considered if plt.pt = FALSE
.
Marc Vandemeulebroecke
Bauer, P., Koehne, K. (1994). Evaluation of experiments with adaptive interim analyses. Biometrics 50, 1029-1041.
Lehmacher, W., Wassmer, G. (1999). Adaptive sample size calculations in group sequential trials. Biometrics 55, 1286-1290.
Vandemeulebroecke, M. (2006). An investigation of two-stage tests. Statistica Sinica 16, 933-951.
adaptTest
package description, CEF
, plotCEF
, tsT
## Compare the tests by Bauer and Koehne (1994), Lehmacher and Wassmer (1999) and Vandemeulebroecke (2006) oldmfcol <- par(mfcol=c(1,3)) pathCEF(typ="b", main="BK 94") pathCEF(typ="l", main="LW 99") pathCEF(typ="v", main="V 06") par(oldmfcol)