You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When normalizing for a truncated Levy, you're dividing by a constant. So maybe memoize it.
I think this is a strategy: use ParetoDistribution's probability(double, double) method (or its deprecated name cumulativeProbability(double double)) to calculate the normalizing constant, i.e. (.probability dist k l) where k is the minimum (probably 1), and l is the max value where the truncation takes place.
Then I think I can just use cumulativeProbability(x), and divide that by the normalizing constant. Since the data will never go above whatever is (treated as) the truncation value, it doesn't matter that the code thinks that a single double passed to cumulativeProbability is from a broader distribution. Or does it?
The text was updated successfully, but these errors were encountered:
When normalizing for a truncated Levy, you're dividing by a constant. So maybe memoize it.
I think this is a strategy: use
ParetoDistribution
'sprobability(double, double)
method (or its deprecated namecumulativeProbability(double double)
) to calculate the normalizing constant, i.e.(.probability dist k l)
where k is the minimum (probably 1), and l is the max value where the truncation takes place.Then I think I can just use
cumulativeProbability(x)
, and divide that by the normalizing constant. Since the data will never go above whatever is (treated as) the truncation value, it doesn't matter that the code thinks that a single double passed tocumulativeProbability
is from a broader distribution. Or does it?The text was updated successfully, but these errors were encountered: