The options $DATA IGNORE=(list) and $DATA ACCEPT=(list) include and exclude selection criteria in the NONMEM data set, respectively. The list has format:

1

 cond,cond,cond,...

where each cond is a test of a data item "label operator value". Comma , is equivalent to .OR. operator, and the .AND. or .NOT. operators are not permitted. One may express .AND. such as in

1

   IGNORE=(GEN.EQ.1.AND.AGE.GT.60)

by negating both sides:

1
2

  .NOT.IGNORE =.NOT.(GEN.EQ.1.AND.AGE.GT.60)
   ACCEPT=.NOT.GEN.EQ.1.OR..NOT.AGE.GT.60

thus

1

   ACCEPT=(GEN.NE.1,AGE.LE.60)

But logical operators inside parenthesis are not permitted, so that the following cannot be coded:

1

 $DATA ACCEPT=((A==1.OR.A==2).AND.B<100)

Even if negated, this is:

1

 IGNORE=(.NOT.(A==1.OR.A==2)).OR..NOT.B<100)

which is:

1

 IGNORE=((A.NE.1.AND.A.NE.2).OR.B>=100)

There is still an .AND. inside parentheses.

There is a workaround that takes advantage of the fact that A takes numeric values and there are no values between 1 and 2.

1

 (A.NE.1.AND.A.NE.2)

can be implemented with an "or":

1

 (A.LT.1.OR.A.GT.2).

The following are equivalent:

1
2

 IGNORE=(B.GE.100)  IGNORE=(A.LT.1) IGNORE=(A.GT.2)
 IGNORE=(B.GE.100,A.LT.1,A.GT.2)

This workaround cannot handle continuous A or more complicated conditions.

With NONMEM through 7.3, two separate NONMEM problems can be used, in two separate runs, for a complicated condition. In the first problem, abbreviated code is used to append a new variable to the data with a value indicating ignore/accept; in the second run the variable is used in the $DATA statement to choose from the data set.

Suppose ACC is the variable that is to have values 0/1 for ignore/accept. The desired code is:

1
2

 ACC=0
 IF ((A == 1.OR.A == 2).AND.B<100) ACC=1

The above can be coded using the FORTRAN language. But NM-TRAN abbreviated code is a subset of FORTRAN. Logical operators .NOT., .AND., .OR. may not be used within parentheses. There are two workarounds.

One is to clear the parentheses:

1

 IF (A==1.AND.B<100.OR.A==2.AND.B<100) ACC=1

this is always possible, no matter how complicted the conditional expression.

The second is to use several statements. There are may ways to do this. In the following, the .AND. of multiple conditions is false if any of the conditions is false.

1
2
3

 ACC=1
 IF (A.NE.1.AND.A.NE.2) ACC=0
 IF (B.GE.100) ACC=0

In prob1.ctl, the first problem appends ACC to the (new) data set. In prob2.ctl, the $DATA record uses ACC to select records.

All required data items for NONMEM (and PREDPP, if used) must be present in the original data file testab.dat. The control stream for the first run:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18

 $PROB test of accept Run #1
  ; Example of implementation of
  ; $DATA ACCEPT=((A == 1.OR.A == 2).AND.B<100)
  $INPUT      NO A B ID DV MDV
  $DATA       testab.dat IGNORE C
  $PRED
  ; Implements:
  ; ACC=1 IF ((A == 1.OR.A == 2).AND.B < 100)
     ACC=1
     IF (A.NE.1.AND.A.NE.2) ACC=0
     IF (B.GE.100) ACC=0
  ; The model for this run is unimportant. Keep it simple.
     Y=THETA(1)+ETA(1)+EPS(1)
  $THETA  1
  $OMEGA 1
  $SIGMA  .4
  $TABLE NO A B ID DV MDV ACC
   NOPRINT NOAPPEND NOHEADER FILE=run1.tab

The control stream for the second run:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14

  $PROB test of accept Run #2
  $INPUT NO A B ID DV MDV ACC
  ; Must do a numeric test for ACC==1, because it appears in
  ; table file run1.tab as 1.0000E+00
  $DATA run1.tab ACCEPT=(ACC.EQN.1)
  $PRED
  ; the model in this problem is the model
  ; used for simulation, analysis, etc.
   Y=THETA(1)+ETA(1)+EPS(1)
  $THETA 1
  $OMEGA 1
  $SIGMA .4
  $TABLE NO A B ID DV MDV
  NOPRINT NOAPPEND NOHEADER FILE=run2.tab

With NONMEM 7.4, it is possible to combine both problems into one NONMEM run using the $TABLE EXCLUDE_BY list option. If any variable in list is non-zero, the record is excluded from the table file. In that case the $DATA ACCEPT/IGNORE option is not needed. Because the table does not exist when NM-TRAN processes the second problem in the control stream, $DATA .. NOOPEN is needed and a format specification is needed.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26

  $PROB  test of NONMEM 7.4 exclude_by feature
  $INPUT      NO A B ID DV MDV
  $DATA       testab.dat IGNORE C
  $PRED
  ; sets ACC=0 if (a==1.or.a==2).and.b<100)
  ; uses the clear the parenthesis approach.
  ACC=1
  IF (A==1.AND.B<100.OR.A==2.AND.B<100) ACC=0
  ; The following is the model for both problems.
  ; It may be more complicated than this.
     Y=THETA(1)+ETA(1)+EPS(1)
  $THETA  1
  $OMEGA 1
  $SIGMA  .4
  ; probe.tab has only those records that have ACC=0.
  $TABLE NO A B ID  DV MDV EXCLUDE_BY ACC
   NOPRINT NOAPPEND NOHEADER FILE=probe.tab

  $PROB test of exclude_by #2
  $INPUT NO A B ID DV MDV
  $DATA probe.tab (10F12.0) NOOPEN
  $THETA 1
  $OMEGA 1
  $SIGMA .4
  ; This problem should have $ESTIMATION and other tasks
 $TABLE NO A B ID DV NOAPPEND NOHEADER NOPRINT FILE=probe2.tab

Here is a fragment of the data testab.dat

 1
 2
 3
 4
 5
 6
 7
 8
 9
10

 C NO A  B   ID DV MDV
   1  1  100 1  0  0
   2  2  100 1  0  0
   3  3  100 1  0  0
   4  1  101 1  0  0
   5  2  101 1  0  0
   6  3  101 1  0  0
   7  1  99  1  0  0
   8  2  99  1  0  0
   9  3  99  1  0  0