Principles of bootstrapping

Bootstrapping is, in some sense, the opposite of sampling from a population. Sampling treats your dataset as the population, and you generate a random subset. Bootstrapping treats your dataset as a sample and uses it to build up a theoretical population.