Write a function in Matlab to compute e^x. Your function should implement a truncated version of the innite sum e^x = lim_n rightarrow infinity sigma_k=0^n x^k/k!. Compare your function with the built-in function exp(x) for x = plusminus1, plusminus 10, plusminus 15. Specifically, for each x, how many terms n do you need to achieve a relative error smaller than 10^-8? Are the results for x > 0 different from those for x