I am writing a multi-equation program using the -ml- commands (sort of
like -mlogit-). I would like to be able to use the value labels of the
dependent variable as the equation names. However, this is proving to be a
non-trivial matter, since value labels need not be valid equation names,
e.g. equation names cannot contain spaces and lots of other characters that
are legit for value labels. Also, value labels need not all be different
for a variable.
I imagine I am not the first person to ever encounter this problem. If
anybody can point me to do some code or other materials that show good ways
to handle this, I would appreciate it.
