3.2.9 The

(Group/Row-Wise)

The `GROUPS`, `ROWS` and `CONSTRAINTS` indicator cards
are used interchangeably to announce the names of the groups
which make up
the objective function
and, for constrained
problems, the names of the
constraints (or rows, as they are often known in linear programming
applications).
In addition, the entries for the linear elements
are input here. The user may give a scaling factor for the groups or
constraints.
Furthermore, groups which are linear combinations of
previous groups may be specified. The syntax for the data cards
which follow these indicator cards
is given in Figure 3.10.

The one- or two-character string in data field 1 specifies the type of group, row or constraint to be input. Possible values for the first character and their interpretations are exactly as in Section 3.2.6.

The string `group-name` in data field 2 gives the name of the
group (or row or constraint)
under consideration. This name may be up to ten characters long
excepting that the name ``SCALE'`
is not allowed. For `X` and
`Z` data cards,
the expanded array name
must be valid and the integer indices must have been defined in a
parameter assignment (see Section 3.2.3). The kind of group
(`N`, `L`, `G` or `E`)
will be taken to be that which is defined on the *first*
occurrence of a data card
for that group.
Subsequent contradictory information will be ignored.

The string `$$$$$$$$$$` in data field 3 is used for
three purposes.

- It may be used to specify that the group mentioned in field 2 has a
linear element
involving the variable given in field 3. In this case,
the string in field 3 must have been defined in the
`VARIABLES`section. If an array definition is being made, the string in field 3 must be an array name. The numerical value of the coefficient of the linear term corresponding to the variable must now be specified. On`Z`cards, the value is that previously associated with the real parameter`r-p-a-name`given in field 5. On other cards, the actual numerical value`numerical-vl`may occupy up to 12 characters in data field 4. - It may be used to announce that all the entries (if any) in the linear
element
for the group under consideration are to be scaled; in this
case field 3 will contain the string
``SCALE'`. The numerical value of the scale factor, that is the factor by which the group is to be divided, is now specified exactly as above.In these first two cases, fields 5 and 6 may be used to define further coefficients or a scale factor for non

`Z`cards. - If the first character in field 1 is a D, the current group is to be
formed as a linear combination of the groups mentioned in fields 3 and
5; the multiplication factors are then recorded in fields 4 and 6
respectively. Thus we will have

In this case, the names of the groups in fields 3 and 5 must have already been defined. The multiplication factors may occupy up to 12 locations in fields 4 and 6.