If $X$ is uniformly $(0,1)$ distributed, how are we sure that $\log X$ is well defined?

90 Views Asked by At

My book says that if $X$ is uniformly (0,1) distributed, then $-\log(X)$ has an exponential distribution with parameter $1$.

However, how are we sure that $- \log(X)$ is well defined?

I.e., can't it be the case that $X$ takes negative values?