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

Should avoid hash table creation in monotonic DatetimeIndex? #1160

Closed
wesm opened this Issue Apr 29, 2012 · 0 comments

Comments

Projects
None yet
1 participant
@wesm
Member

wesm commented Apr 29, 2012

In very large time series, a first call to

ts[stamp]

will cause a hash table population to look up the location of stamp. If the index has, say, 10mm timestamps, the allocated hash table will take up at minimum 160 megabytes ((timestamp, location) pairs, each 8 bytes) probably rounded up to 256 megabytes. I would say either always use binary search or don't create the hash table when the index exceeds a certain length.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment