MODULOPT
Optimization Routines
Naming convention
The naming convention aims at condensing the different features of an
optimization routine in its name.
The name of a routine has the following structure:
where
-
M
is the letter M or N: M for a single precision code, N for a double
precision code; this convention has a tendency to change, since all
codes should now be written in double precision and all start with the
letter M;
-
x
is a number specifying the type of optimization problem, the code can
solve:
-
1: unconstrained problems,
-
2: bound constrained problems,
-
3: linearly constrained problems,
-
4: problems with general nonlinear constraints;
-
TT
are two letters describing the type of algorithm implemented in the
code;
-
y
is a number giving the version of the code with the same root MxTT.
Some routines have an easierly memorizable nickname.