Skip to content
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.

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

Fixed median checknan help message #8605

Closed
wants to merge 3 commits into from
Closed

Fixed median checknan help message #8605

wants to merge 3 commits into from

Conversation

tinybike
Copy link

@tinybike tinybike commented Oct 7, 2014

No description provided.

@ivarne
Copy link
Member

ivarne commented Oct 7, 2014

We should probably be clare on what happens if checknan=false also?

Also reference the first PR #8602

@simonster
Copy link
Member

Does the checknan option even make sense? It doesn't ignore the NaNs; it just gives the median as if they're sorted to the right. Is there a reason someone would want that behavior?

Also, the change to base.rst needs to be in restructured text format, i.e. \" should be double backticks. If you want to update helpdb.jl you should do so by running make -C doc helpdb.jl, although you can also just change base.rst and wait for someone else to update helpdb.jl.

@simonster
Copy link
Member

cc @StefanKarpinski since it looks like you first implemented checknan. Is this just a performance optimization to avoid having to check the entire vector for NaNs?

@StefanKarpinski
Copy link
Member

Yikes. I don't even remember doing that. But yes, I think it's just a performance optimization.

@simonster
Copy link
Member

If I add @inbounds, the cost of the NaN check is about 4%. Is that worth the extra option?

@StefanKarpinski
Copy link
Member

Probably not. Just get rid of it.

@simonster simonster closed this in e24fac0 Oct 7, 2014
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants