+--------------------------------------------------------------------+
 |                                                                    |
 |                        DATA_AVERAGE EXAMPLE                        |
 |                                                                    |
 +--------------------------------------------------------------------+

 A  template record is a data record that contains a non-zero value for
 the raw-data (RAW_) data item.  Only the DV values with those observa-
 tion  records  (i.e.,  records  with MDV=0) matching the template data
 record will be included in the average.  This average replaces the  DV
 item  in  the  template  record, and it may be displayed in a table or
 scatterplot.

 An observation record matches a template if all  its  data  items  are
 identical to all those on the template, except that the values of NON-
 MEM data items do not affect the match.  (Any nonNONMEM data item type
 may in fact be excluded from the match (See omit)).

 Example 1

 Here is a fragment of a data set (which may or may not actually be for
 PREDPP):

 #ID TIME AMT WT AGE EVID MDV DV RAW_ GEND
  1  5    0   70 50  0    1   2  1     1   template record
  4  5    0   70 50  0    0   4  0     1   observation record

 These two records match.  The NONMEM items are the ID,  MDV,  DV,  and
 RAW_  items.  The DV item with the observation record will be included
 in the average.  The DV item in the template record is ignored for the
 purpose  of  computing the average (MDV=1); the displayed DV item with
 this record will be the average value, and the displayed RES item will
 be the difference between this average and the PRED item.  The average
 is taken over all records matching the TIME, AMT, WT, AGE,  EVID,  and
 GEND items of the template record.

 If the omit record

 $OMIT GEND

 appears, then the two records

 #ID TIME AMT WT AGE EVID MDV DV RAW_ GEND
  1  5    0   70 50  0    1   2  1     1   template record
  4  5    0   70 50  0    0   4  0     2   observation record

 also match.  In this case the average is taken over all records match-
 ing the TIME, AMT, WT, AGE, and EVID items of the template record.

 Example 2

 Here is an example of categorical dose-response  with  four  different
 bolus  doses per individual (separated by washouts), and data averages
 are formed for each of the four doses.

 #ID DOSE DV MDV RAW_
  100 5    .  1   1    set DV item to avr. for DOSE=5
  100 10   .  1   1    set DV item to avr. for DOSE=10
  100 20   .  1   1    set DV item to avr. for DOSE=20
  100 40   .  1   1    set DV item to avr. for DOSE=40
  1   5    1  0   0
  1   10   1  0   0
  1   20   2  0   0
  1   40   2  0   0
  2   5    1  0   0
  2   10   1  0   0
  2   20   1  0   0
  2   40   1  0   0
  3   5    1  0   0
  3   10   2  0   0
  3   20   2  0   0
  3   40   3  0   0
  .
  .

 Example 3

 An average of a PRED-defined item may also be displayed.

 $ABBREV COMRES=1 COMSAV=1
  ...
 $PRED
  COM(1)=LOG(DV)
  ...

 $TABLE COM(1)=LDV

 For template records, the value of LDV displayed in the table  is  the
 average of LOG DV over all observation records matching the template.

 (See data average block, raw).

 REFERENCES: none.


  
Go to main index.
  
Created by nmhelp2html v. 1.0 written by Niclas Jonsson (Modified by AJB 5/2006,11/2007,10/2012)