Join GitHub today
GitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.
Sign upGitHub is home to over 50 million developers working together to host and review code, manage projects, and build software together.
Sign up
Dear Alboukadel KASSAMBARA,
Thanks for a great library. I have been searching for a library to calculate a confidence interval of Cohen’s d from a Welch test. And the way you implement bootstrapping to do that seems like a way to go.
However, when I use the cohens_d function, it returns only positive values for effsize, conf.low, conf.high. So I guess, the effsize is the absolute value of d, or |d| and the conf values are also the absolute values of the bootstrapping bounds. This is problematic when conf.low and conf.high have different directionality (e.g., -.2 to .3). Is there a way to output values with directionality, as opposed to using absoluate values?
Thanks so much,
Narun