Optional mask for jacobian evaluations in JMI


For every Jacobian evaluation (in PyJMI) one must supply a list of
independent variables and a mask of which Jacobian columns that are to
be evaluated. This is somehow counter intuitive as the variable
independent_variables in a way defines a mask and is also
easier to work with.

This patch proposes making the mask parameter optional
for all JMI functions that evaluate Jacobians. This makes the JMI
Jacobian interface functions more intuitive and easier to use for less
experienced users. It also makes it easier to work with smaller models
where the mask parameter is not as (or at all) important.

If this patch is applied the PyJMI interface will be slightly
different from the JMI interface.