.-
help for ^tabby^              Version 2/10/96
.-

Tabulation with more control over format
----------------------------------------

 ^tabby^ varlist  [^if^ exp] [^in^ range] [^,^ ^b^y(var)  ^b^ymis ^m^is 

  ^d^ec^(^integer^)^ ^w^idth^(^#^)^ ^h^ead ^t^otal ^p^erc|^c^ol|^r^ow ]

  no^byt^ot  ^s^tart^(^int 10^)^  no^z^ero ^max(^int 100^)^ ^te^st^(^string^)^


Generates a table with frequencies or row or column percentages of the variable
> s
in ^varlist^. If ^by(var)^ is specified, a table of crosstabulations of
each variable in the varlist with the by-var is generated.

String variables in the varlist or as byvar are trimmed of any leading or 
trailing blanks before analysis. Empty strings are considered missing. 

Options
-------
 ^by(^var^)^  Optional. If not specified only overall oneway frequencies or
    column percentages are shown.

 ^dec(^..^)^    Specifies number of decimals for percentages. Default 0.
    Frequencies are always displayed without decimals.
    If the length of the statistic would exceed the available 
    width for the display, the number of decimals is reduced.

 ^width^  Integer number, with default value 10. Specifies the minimum width 
    for each column. The value 10 corresponds to the width of columns created
    by tabulate. A specified width is adapted to the minimum number of digits
    required to display a percentage (3 digits) with decimals, if specified, 
    and the total number of observations. The width excludes a space between 
    columns. If the specified width is too small for a column, the width will 
    be increased for that column.
    Long column-labels in the header are abbreviated if necessary.

 ^mis^    Specifies that the category with missing values of each
    variable in the ^varlist^ should be included in the table.
    Default the missing category is excluded. 
    For string variables an empty string or a string consisting
    only of blanks is considered missing.

 ^bymis^  Specifies that the category with missing values of the ^byvar^
    should be included in the table. Default these are excluded.

 ^head^   Request the display of a heading of the table. Active only if
     ^by(var)^ is specified.
 ^total^        Request the display of a row(s) with totals for each variable
    in the ^varlist^ if ^mis^ is not specified. If ^mis^ is 
    specified, one Total -row will be displayed at the top of 
    the Table.

 ^row^    Request row-percentages. Only in case a byvar is specified  
    with at least two non-empty categories. In the total column 
    of the Table frequencies will be displayed.

 ^col^    Request column-percentages. In the Total-row(s) of the Table
    frequencies will be displayed.

 ^perc^   Requests column-percentages in ^addition^ to the frequencies if
           ^by()^ is not specified. Otherwise equivalent to ^col^. 

 ^head^   Request the display of a heading of the table. Active only if
     ^by(var)^ is specified.

 ^nobytot^    Suppresses the output of a total-column. Default a total-column 
    is printed. Option is ignored if row-percentage is requested.

 ^start()^  Specifies the (minimum) position on a line for the start of
    the data-columns, i.e. the column in which the ^|^ is printed.
    Default and minimum 12. May be increased by the program if 
    long string variables are in the varlist.

 ^nozero^ Requests for display of a "." in cells with zero-frequencies.

 ^max()^  Specifies the maximum number of rows per variable, or number
    of different unique values of a variable from the varlist 
    that is allowed. Default 100. In case a variable has more 
    than ^max^ different values, the variable label will be 
    displayed together with a message. This option is useful to 
    prevent long tables. 
    
 ^test()^ request the calculation of a P-value:
  test(^chi^square) : Pearson Chisquare test (see note below)
  test(^kw^allis)   : Kruskal-Wallis by ^by^-variable
  test(np^tr^end)   : NPtrend test by ^ordinal by^-variable
  test(^sp^earman)  : Spearman rank test    
  test(^rh^o)       : Spearman rank test    
  test(^exac^t)     : Fisher's exact test
  Note: only non-missing observations on both the row and column variable
        are included in the calculation of the P-value independent of the
        specification of ^mis^ or ^bymis^.
  ^Note on Chisquare test:^ this test will ^not^ be performed if in any column
  (resp. row) the column- (resp. row-)total divided by the number of rows 
  (resp. columns) is less than 3.  
  
        
 ^Examples:^
 ^---------^

. tabby t erpr, by(nn) mis head
. tabby t erpr, by(nn) mis head  bymis col dec(1) nobytot
. tabby t erpr, by(nn) mis wi(5) bymis row dec(1) test(chi)
. tabby t erpr,  mis wi(5)  col dec(1)

^Author:^
^-------^ 
  Wim van Putten
  Erasmus MC - Daniel den Hoed Cancer Center
  Rotterdam
  The Netherlands
  FAX: +31.10.4391028
  e-mail: w.vanputten@@erasmusmc.nl