Next: fromclip() Up: MacAnova Help File Previous: fprint()   Contents


result <- formatpval(pvals [,minpval:minp] [,format:fmt]), non-negative
  REAL vector or scalar, nonnegative REAL scalar minp <= .001, CHARACTER
  scalar fmt

Keywords: output
formatpval() is designed to be used in a macro to format printed
P-values similarly to the way the built-in GLM functions such as
regress(), anova() and logistic() do.

result <- formatpval(pvals), where pvals is a REAL scalar or vector with
min(pvals) >= 0, computes the CHARACTER scalar or vector result the same
length as pvals.

If pvals[i] < minP, where minP is the value of option 'minpval', then
result[i] has the form, for example, "< 1e-8", when minP = 1e-8.  If
pvals[i] >= minP, result[i] is simply paste(pvals[i]).

result <- formatpval(pvals, minpval:minP), where 0 <= minP <= .001 does
the same, except the supplied value of minP is used.

result <- formatpval(pvals, format:fmt [, minpval:minP]), where fmt is a
CHARACTER scalar specifying a legal output format (see subtopic
options:"format"), does the same, except fmt is used to format result,
instead of the default format.

  Cmd> tstats <- vector(4.585, 5.536, 6.576, 7.611, 8.826)

  Cmd> formatpval(twotailt(tstats,23),minpval:1e-7,format:".3g")
  (1) "0.000131"
  (2) "1.25e-05"
  (3) "1.04e-06"
  (4) "< 1e-07"
  (5) "< 1e-07"

See also print(), write(), options:"minpval", options:"format",

Gary Oehlert 2003-01-15