8000 ndcg different from ndcg_cut_20 for 20 results ? · Issue #11 · usnistgov/trec_eval · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

ndcg different from ndcg_cut_20 for 20 results ? #11

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
gaurav5430 opened this issue Apr 11, 2017 · 2 comments
Open

ndcg different from ndcg_cut_20 for 20 results ? #11

gaurav5430 opened this issue Apr 11, 2017 · 2 comments

Comments

@gaurav5430
Copy link

I have not set the relevance gains level for ndcg from the command line.
So, everything is at default, and i have 20 results per query, then why is ndcg_cut_20 different from ndcg ?
(although i can see that they are same after ndcg_cut_100) , am i missing something?
@gaurav5430
Copy link
Author

It seems that for multiple queries, it is taking the maximum of num_rel across these queries and then calculating ndcg@max(num_rel) for all of them.

For example :

Q1 has 20 num_ret and 30 num_rel
Q2 has 20 num_ret and 35 num_rel
Q3 has 20 num_ret and 28 num_rel

then ndcg returns me a score which is same as ndcg_cut_35
while i expect it to give me ndcg@20, which is the number of results for each query.

@gaurav5430
Copy link
Author

for Q1 : ndcg = ndcg_cut_30
Q2 : ndcg = ndcg_cut_35
Q3 : ndcg = ndcg_cut_28

for all ndcg = ndcg_cut_35

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@gaurav5430
0