Next: options
Up: MacAnova Help File
Previous: NULL
Contents
Keywords:
variables, syntax, missing values
You enter numbers as integers with or without out a decimal point, as
decimal numbers, or using exponential notation (X.XXeY or X.XXEY is XX.X
* 10^Y).
Cmd> a <- 65535 # same as a <- 65535. or a <- 65535.0000
Cmd> a <- -3141.592654; b <- .0000415; c <- 1000000
Cmd> a <- -3.141592654e3; b <- 4.15E-5; c <- 1e6# same as preceding
For greater readability of long numbers, you can use '_' to separate
digits in the mantissa (the entire number without 'e' or 'E' or the part
before 'e' or 'E').
Cmd> a <- -3_141.592_654e4 # same as a <- -3141.592654e4
Warning: _3_141, for example, is a legal variable name, not a number.
Fortran style double precision numbers like -3.14592654d3 and 4.15D-5
are read as if the 'd' or 'D' were 'e' (except by read() and matread()).
It is an error to attempt to enter a number that is too large to be
represented in the computer. For example,
Cmd> d <- 3.1e5000
is an error. On most computers the largest numbers are about +-2^1024 =
+-1.79769e+308 and the smallest nonzero numbers are about +-2^(-1024) =
5.56268e-309.
You enter MISSING values using the symbol '?'.
Cmd> e <- vector(1,2,3,?,4,5,?) # vector with 2 MISSING values.
Other representations for MISSING such as '*', 'NA' and '.' which are
recognized by vecread() are not recognized in MacAnova commands.
By default, most numeric output is printed using a hybrid between
integer, decimal and exponential format. MISSING values are normally
printed as 'MISSING'.
Cmd> vector(100*PI, 1e6*PI, PI/100, PI/1e6)
(1) 314.16 3.1416e+06 0.031416 3.1416e-06
Cmd> vector(34, ?)
(1) 34 MISSING
You can change these defaults by setoptions() keywords 'format' and
'missing'. Here is an example:
Cmd> setoptions(missing:"NA", format:"12.4f")
Cmd> vector(100*PI, 1e6*PI, PI/100, PI/1e6)
(1) 314.1593 3141592.6536 0.0314 0.0000
Cmd> vector(34, ?)
(1) 34.0000 NA
See topics setoptions(), 'options', print().
See also topic 'syntax'.
Gary Oehlert
2003-01-15