ST500                               Session 4 _SPSS:  set length=59
width=80 {this may be inserted at any place in the control file
        and is in effect until changed again}
_SAS:  options ls=80 ps=59; {this may be inserted before a data step,
or with a proc
        and is in effect until changed again}
_review assignment 1 (a) sas - output _pc - statistics programs: American
Statistician  + PC Magazine

_minitab: syntax + commands * minitab + sas were the easiest to change
to interactive pgms, spss required major syntax changes.  * actually
better to run minitab as interactive
  journal='file' is a way to capture your interactive session in a file
  which you can later edit (with
pico or vi) and then run in batch
 _remember: man minitab>f1 to get syntax in a file to print

_If Minitab requires more memory, include 'setenv MINISIZE 300000' in
your .login file (100000 is default size). The statement is located in my
.login shown in the login.example we looked at earlier and can be found
under netscape: Eamples/Handouts/login.example .  _ minitab <control
{batch mode} creates control.LIS (need echo/outfile/noterm/nooutfile)
   or minitab {interactive mode}

                           Assignment 1 (c) - minitab

This assignment is due tuesday, jan 9 - please email me the listing of
the output (a1c.LIS) Note that many minitab commands can not use alpha
variables (i.e. sex='m' or 'f'), so make a copy of the assign1.dat file
and use pico (vi) to change all 'm' to '1' and all 'f' to '2'. You might
do it in the following manner:
        cp assign1.dat assign1.datx
    pico assign1.datx
        {Then scroll down the sex column and replace 'm' with '1' and
        'f' with '2'}

The assignment is the same series of 10 operations that you did on the
data file 'assign1.dat' with sas and spss, but now use 'assign1.datx'.

To help you out a 'little' bit, the statements are listed on the next
page!  Note that if you name columns, you need the single quotes around
the name (i.e. 'sex') every time you use the name.  Also if a 'command'
has a 'subcommand' or multiple 'subcommands', that
        1)each 'command' / 'subcommand' MUST be on separate lines 2)each
        'command' / 'subcommand' (except the last one) MUST have a
        ';' {semi-colon} at the end, and
    3) the last 'subcommand' must have a '.' {period} at the end.
        An example: sort c1 into c4;
                by c2.
note: to run in batch:        %minitab <a1c  (creates a1c.LIS with the
output) note: to run interactively:   %minitab and type all  lines except
echo/outfile/noterm/nooutfile lines echo outfile 'a1c';
 noterm.
read 'assign1.datx' c1-c5;
  format( a10,1x,f1,f5.1,f6.1,f3.0).
name c1='name' c2='sex' c3='height' c4='weight' c5='age' code (-1) to * in
'weight' 'weight' print 'name'-'age' tally 'height'-'age' table 'sex';
  counts; means 'height'-'age'.
unstack ('name'-'age') (c11 - c15) (c21 - c25);
  subscripts 'sex'.
print c11-c15 c21 -c25 stack (c11-c15) (c21-c25) in (c31-c35) print c31
- c35 sort 'name'-'age' 'name' - 'age';
  by 'sex' 'name'.
lplot 'weight' 'height' letter 'sex' lplot 'weight' 'age'          'sex'
logten 'weight' c6 name c6='lwt' lplot 'lwt' 'age' 'sex' stem-and-leaf  c4
stem-and-leaf  c5 stem-and-leaf  c6 boxplot 'weight' boxplot 'height'
boxplot 'age' boxplot 'lwt' histogram 'weight' histogram 'height'
histogram 'age' histogram 'lwt' random 200 obs in c41;
  normal.
mult c41 .2 c42 add 1 to c42 put c42 mult c42 c42 put c43 raise 10
c42 put c44 name c41='a' c42='b' c43='c' c44='d' boxplot 'a' boxplot
'b' boxplot 'c' boxplot 'd' histogram 'a' histogram 'b' histogram 'c'
histogram 'd' nooutfile stop