Syntax
where B and
The first form of the syntax defines and prepares the base B for
adjustment, but does not carry out any adjustment. Thus, for example,
the following two sequences of commands are equivalent:
where N1, N2 are the names of bases. The SCAN: and
ITADJUST: commands can be used similarly to force preparation
for adjustment. The preparation is the same for any of these commands,
so it doesnt matter which one you use.
The base notation discussed in §9.1.2 may be used. Any base
which is presented for use during an adjustment must contain
only elements having beliefs, and no data elements.
The second form of the syntax is used for explicitly partial
adjustments, and assumes that the base to be partially adjusted is
defined already, either by the current adjustment, or by having prepared
a base B for adjustment using the first form of the syntax. Suppose
that the current adjustment (i.e. before this command is issued) is
Regarding the data used for partial adjustments, subsequent partial
adjustments following an initial adjustment must fit quantities such
that they possess data consonant with the data selection used for the
initial adjustment. Data averages are calculated afresh for every
element in the supplied partial list, so it is possible to carry out the
adjustment
The preparation of adjustment is most useful
when a sequence of adjustments is envisaged, as in the following
example:
are the names of bases or
elements.
. Then the following sequence of adjustments is carried out.
Firstly, the base B is adjusted by the base 
, giving
the adjustment
. Secondly, the base 
is adjusted by the base
. At each stage both the overall adjustment and
the partial adjustment is reported. That is, we report both and 
, where the latter is the partial
adjustment of B by 
, having accounted for 
. This
process continues for the remaining information sources 
: at
each stage, we report the overall adjustment to that point, together
with the
corresponding partial adjustment representing the change between the new
and former adjustments.
, change the data on E, carry out the adjustment
and so obtain different adjustments.
+ D 
+ ...+ D 
]