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

Updated pwr.p.test #4

Merged
merged 4 commits into from
Oct 6, 2016
Merged

Updated pwr.p.test #4

merged 4 commits into from
Oct 6, 2016

Conversation

aanandku
Copy link

Changed the upper limit of n from 5 to 9 digits, for the function to be able to work in situations where n is greater than or equal to 100,000 (lines 52-53). I verified that the program didn't take any longer than it did when the upper limit was 5 digits. For internet-scale testing, it's frequently the case that the effect size is very small, and the number of observations required is larger than 100,000. For example, for p0=5% and p1=5.15%, the number of observations required is 132367. Small effects like this over large populations are especially common in e-commerce, search, and advertising. The legacy function returns an error in such situations because the search space for n is limited to 100,000. This change increases the search interval for sample size to 10 million, to account for these sorts of scenarios.

heliosdrm and others added 2 commits September 5, 2016 16:01
Changed the upper limit of n from 5 to 9 digits, for the function to be able to work in situations where n is greater than or equal to 100,000 (lines 52-53). I verified that the program didn't take any longer than it did when the upper limit was 5 digits. For internet-scale testing, it's frequently the case that the effect size is very small, and the number of observations required is larger than 100,000.  For example, for p0=5% and p1=5.15%, the number of observations required is 132367.  Small effects like this over large populations are especially common in e-commerce, search, and advertising.  The legacy function returns an error in such situations because the search space for n is limited to 100,000.  This change increases the search interval for sample size to 10 million, to account for these sorts of scenarios.
@heliosdrm
Copy link
Owner

Thanks for the contribution. Would you be willing to revise the upper limit of n for the rest of functions? If such change is done, it would be advisable to apply it coherently to the whole package, not only in a particular function.

Increased the limit for n
increased limit for n
@aanandku
Copy link
Author

Am working on changing the upper limit of n for the rest of the functions at the moment

@heliosdrm heliosdrm changed the base branch from master to workshop October 6, 2016 21:11
@heliosdrm heliosdrm merged commit e651836 into heliosdrm:workshop Oct 6, 2016
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.

None yet

3 participants