Open
Description
Hi William,
I have a question regarding the output of cossim. I use two string variables as the input: x = "111111111", and y = "011010001". The output from the "strdist x y, cossim(cossim)" is 0.75592895.
If I treat each of the string variables as a vector of nine digits, and use the cosine similarity formula (x . y) / (||x|| . ||y||), the value is 2 / 4.2426 = 0.4714
I am wondering what causes the discrepancies here?
Best,
Henry
Metadata
Metadata
Assignees
Labels
No labels