The AMS math packages define a number of useful math alignment structures. These have been well thought out and designed with particular logical structures in mind, as well as the layout. Thus these environments are less often abused than is eqnarray. In this section, we list the environments, their expected use case and describe the strategy used for converting them.
To be done Describe alternates for equation and things inside equations; Describe single vs multiple logical equations. (and started variants)
This list outlines the intended use of the AMS alignment environments The following constructs are intended as top-level environments, used like equation.
Several of the constructs are used in place of a top-level equation and represent one or more logical equations. The following describes the intended usage, as a guide to understanding the implementation code (or its limitations!)
align,flalign,alignat,xalignat:
Each row may be numbered; has even number of columns;
Each pair of columns, aligned right then left, represents a logical equation;
Note that the documentation suggests that annotative text can be added
by putting \text{} in a column followed by an empty column.
gather: Each row is a single centered column representing an equation.
multline: This environment represents a single equation broken to multiple lines; the lines are aligned left, center (repeated) and finally, right. alignment not yet implemented
The following environments are used within an equation (or similar) environment and thus do not generate MathFork structures. Moreover, except for aligned, their semantic intent is less clear. The preservation of the alignment have not yet been implemented; they; presumably would yeiled an XMDual.
split
gathered
aligned,alignedat
Note that the case of a single equation containing a single aligned is transformed into and treated equivalently to a top-level align.