/[pspp]/pspp/doc/transformation.texi
ViewVC logotype

Diff of /pspp/doc/transformation.texi

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1.2 by blp, Mon Feb 14 00:09:20 2005 UTC revision 1.3 by blp, Sun Mar 13 07:21:14 2005 UTC
# Line 22  as a rule. Line 22  as a rule.
22  @vindex AGGREGATE  @vindex AGGREGATE
23    
24  @display  @display
25  AGGREGATE  AGGREGATE
26          /BREAK=var_list          OUTFILE=@{*,'filename'@}          
27          /PRESORTED          /PRESORTED
         /OUTFILE=@{*,'filename'@}  
28          /DOCUMENT          /DOCUMENT
29          /MISSING=COLUMNWISE          /MISSING=COLUMNWISE
30            /BREAK=var_list
31          /dest_var['label']@dots{}=agr_func(src_vars, args@dots{})@dots{}          /dest_var['label']@dots{}=agr_func(src_vars, args@dots{})@dots{}
32  @end display  @end display
33    
# Line 36  Cases are divided into groups that have Line 36  Cases are divided into groups that have
36  variables called @dfn{break variables}.  Several functions are available  variables called @dfn{break variables}.  Several functions are available
37  for summarizing case contents.  for summarizing case contents.
38    
39  At least one break variable must be specified on BREAK, the only  The OUTFILE subcommand is required and must appear first.  Specify a
40  required subcommand.  The values of these variables are used to divide  system file by file name string or file handle (@pxref{FILE HANDLE}).
41  the active file into groups to be summarized.  In addition, at least  The aggregated cases are written to this file.  If @samp{*} is
42  one @var{dest_var} must be specified.  specified, then the aggregated cases replace the active file.
43    
44  By default, the active file is sorted based on the break variables  By default, the active file will be sorted based on the break variables
45  before aggregation takes place.  If the active file is already sorted  before aggregation takes place.  If the active file is already sorted
46  or otherwise grouped in terms of the break variables, specify  or otherwise grouped in terms of the break variables, specify
47  PRESORTED to save time.  PRESORTED to save time.
48    
 The OUTFILE subcommand specifies a system file by file name string or  
 file handle (@pxref{FILE HANDLE}).  The aggregated cases are written to  
 this file.  If OUTFILE is not specified, or if @samp{*} is specified,  
 then the aggregated cases replace the active file.  
   
49  Specify DOCUMENT to copy the documents from the active file into the  Specify DOCUMENT to copy the documents from the active file into the
50  aggregate file (@pxref{DOCUMENT}).  Otherwise, the aggregate file will  aggregate file (@pxref{DOCUMENT}).  Otherwise, the aggregate file will
51  not contain any documents, even if the aggregate file replaces the  not contain any documents, even if the aggregate file replaces the
52  active file.  active file.
53    
54    Normally, only a single case (for SD and SD., two cases) need be
55    non-missing in each group for the aggregate variable to be
56    non-missing.  Specifying /MISSING=COLUMNWISE inverts this behavior, so
57    that the aggregate variable becomes missing if any aggregated value is
58    missing.
59    
60    If PRESORTED, DOCUMENT, or MISSING are specified, they must appear
61    between OUTFILE and BREAK.
62    
63    At least one break variable must be specified on BREAK, a
64    required subcommand.  The values of these variables are used to divide
65    the active file into groups to be summarized.  In addition, at least
66    one @var{dest_var} must be specified.
67    
68  One or more sets of aggregation variables must be specified.  Each set  One or more sets of aggregation variables must be specified.  Each set
69  comprises a list of aggregation variables, an equals sign (@samp{=}),  comprises a list of aggregation variables, an equals sign (@samp{=}),
70  the name of an aggregation function (see the list below), and a list  the name of an aggregation function (see the list below), and a list
# Line 72  list. Line 81  list.
81  Each set must have exactly as many source variables as aggregation  Each set must have exactly as many source variables as aggregation
82  variables.  Each aggregation variable receives the results of applying  variables.  Each aggregation variable receives the results of applying
83  the specified aggregation function to the corresponding source  the specified aggregation function to the corresponding source
84  variable.  Most aggregation functions may be applied to numeric and  variable.  The MEAN, SD, and SUM aggregation functions may only be
85  short and long string variables.  Others, marked below, are restricted  applied to numeric variables.  All the rest may be applied to numeric
86  to numeric values.  and short and long string variables.
87    
88  The available aggregation functions are as follows:  The available aggregation functions are as follows:
89    
# Line 173  from calculations.  To include user-miss Line 182  from calculations.  To include user-miss
182  (@samp{.}) between the function name and left parenthesis  (@samp{.}) between the function name and left parenthesis
183  (e.g.@: @samp{SUM.}).  (e.g.@: @samp{SUM.}).
184    
 Normally, only a single case (for SD and SD., two cases) need be  
 non-missing in each group for the aggregate variable to be  
 non-missing.  Specifying /MISSING=COLUMNWISE inverts this behavior, so  
 that the aggregate variable becomes missing if any aggregated value is  
 missing.  
   
185  @cmd{AGGREGATE} both ignores and cancels the current @cmd{SPLIT FILE}  @cmd{AGGREGATE} both ignores and cancels the current @cmd{SPLIT FILE}
186  settings (@pxref{SPLIT FILE}).  settings (@pxref{SPLIT FILE}).
187    

Legend:
Removed from v.1.2  
changed lines
  Added in v.1.3

savannah-hackers-public@gnu.org
ViewVC Help
Powered by ViewVC 1.1.26