OTE="John Lee, post: 52286, member: 31"]Because that gives the correct simulated values? Sorry, but your question is a little vague.[/QUOTE] Sir,I juz meant det if for u1<0...we return y=ln(1-u2)....will it be ok Coz we need half of observations positives & half negatives....
And sir..Its given that x is a real number...zero is also a real number...so how we will get zero by this method...