Cross Validated is a question and answer site for people interested in statistics, machine learning, data analysis, data mining, and data visualization. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I'd like to fit some standard distributions, say lognormal, to a set of data which unfortunately has ties. The precision of the measurement simply isn't high enough. Of course you can still fit the data but all goodness-of-fit test like KS and the like throw warnings - which is natural, since this case cannot happen for continuous distributions.

I tried to add some arbitrary small random $\epsilon$ to every data point to remove the ties without altering the data too much. As expected this results in ridiculously low p-values.

On the other hand: The data consists of some measurements per year. Taking only the average value/year one can fit a distribution very well. So I think it's really because of the ties.

Is there any magic solution to this problem?

share|improve this question

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.