I have a monte carlo simulation that samples three independent variables from their respective (continuous and skewed) distribution. I need to fix one variable and only sample from the other two distributions. What fixed value would be the best representative? median?