set.seed（）函数 - 设置种子的最佳值是什么？

Question

I was wondering if a certain seed value was better for a norm dist sample of data as opposed to a set that is skewed. I understand that using the same seed on the same data will get you the same randomly generated records, I just wonder if seed 1-10 was better a neg skewed data range and 90-100 was better on a positive skew.

Answer 1

The only real benefit of a seed is that your results are reproducible. Ranges of seeds (at least as far as I know) aren't likely to be skewed in one direction more than any other, they just provide a concrete seed for the pseudo random generator rather than to use some entropy in the system.