Takes a vector of real, non-negative weights and converts them to an integer
form while maintaining the ratios between all elements.
Usage
normalize_weights(w, len_out = NA)
Arguments
- w
vector of weights
- len_out
positive integer, desired output length of the normed weights
Value
vector of scaled weights, guaranteed to be integer
Details
Iteratively gets the LCM of all the denominators, so we are
guaranteed to get integer results after multiplying by that LCM. And, because
it is the *least* common multiple, these will be the smallest possible
integer weights.