i have few frequency distributions of categorical data. for example

sample 1: [item1: 9 times,  item2: 10times, item3: 17 times, item4:20 times, item 6: 3]
sample 2: [item1: 17 times, item2: 18times, item3: 3 times,  item4:7 times, item 7: 1]
sample 3: [item1: 50 times, item2: 50times, item3: 5 times,  item4:7 times. item 8: 9]

based on the above same we could say that sample 2 and sample 3 are similar to each other while sample 1 is different than sample 2 and 3.

What are the ways/tests i can detect if samples are similar to each other. the categories might be sparse in nature hence we would have to handle missing ids between 2 samples (for example item 6 is only there in sample 1)



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.