control when to use exponential notation in num2str

I would like to control when exponential notation is used vs when it is not for compact exponential notation in matlab.

For example.

num2str([0.0077;0.324;0.0000435],'%2.4G')

gives

    '  0.0077'
    '   0.324'
    '4.35E-05'

but I want it to give

    '  7.7E-03'
    '   0.324'
    '4.35E-05'

I'm aware of using the "E" notation to alway force exponential notation but I do not like the trailing zeros. I want to be able to specify that anything below 10^-2 should use compact exponential format. I suspect that it is an undocumented element of format spec.

Currently I have this nasty work around.

    matlab_workaround=0.0077
        if log10(matlab_workaround)<-2;
            matlab_workaround=num2str(matlab_workaround,'%2.4E');
            while contains(matlab_workaround,'0E')
                matlab_workaround=strrep(matlab_workaround,'0E','E');
            end
        else
            matlab_workaround=num2str(matlab_workaround,'%2.4G');
        end

Answer 1

[Walter Roberson]

There is no control over conversion other than the format specification the way you are passing it in.

Note: num2str() uses sprintf() (or possibly sprintfc()), so it has the same restrictions that sprintf() has -- which is to say that it is not possible to customize the details of %g conversion.