pchasani
commented
2 months ago Hi, I would like to thank you for sharing these routines. Do you happen to share the replication code for your research paper?
When looking at the DEMO.m file, I am confused about how to obtain a p-value for unimodality. Is that enough to say that checking whether the set S is empty is sufficient for the test to say the data is from a multimodal distribution? Meaning that the procedure does not produce p-values. Apologies if I misunderstood something and I appreciate your help in advance
Hi, thanks for your comment.
UU-test doesn't return a p-value for unimodality. It first decides unimodality/multimodality for a data distribution and in the case of unimodality, the set S gives you the uniform mixture model (UMM). You have correctly understood that if the data distribution is multimodal, the set S is empty. However, if you need a p-value for unimodality, Hartigans' dip-test is a choice for you. I hope i have answered your question clearly.
Hi, I would like to thank you for sharing these routines. Do you happen to share the replication code for your research paper?
When looking at the DEMO.m file, I am confused about how to obtain a p-value for unimodality. Is that enough to say that checking whether the set S is empty is sufficient for the test to say the data is from a multimodal distribution? Meaning that the procedure does not produce p-values. Apologies if I misunderstood something and I appreciate your help in advance