Overflow of fractional representations of floating point numbers?

So if I have a matrix of floating point numbers A, I know that I can make it symbolic:

A = sym(A, 'f');

Now, each of the entries is represented by a fraction.

I am concerned with possible overflow. As I perform more and more operations on these entries, the numerator and denominator of each entry becomes larger and larger. So what I am wondering is, what is the range of the numerator and the denominator?

How can I prevent overflow? If a numerator or denominator exceeds its range, how can I retain the preciseness of the fraction?

Thanks

Answer 1

[Walter Roberson]

You are limited to 2^29 + 1 decimal digits it appears to me. You might encounter practical limitations a lot sooner than that, though.