search

GAA-UAM / scikit-fda

Functional Data Analysis Python package
https://fda.readthedocs.io
BSD 3-Clause "New" or "Revised" License
297 stars 53 forks source link

Effect size in functional One-way ANOVA #620

Open alberto-jj opened 2 months ago

alberto-jj commented 2 months ago

Motivation

Hi, I have been testing the functional one-way ANOVA and also checked the Cuevas A, et al,. paper. However, I haven't seen anything regarding effect size.

Desired functionality

Can I calculate any standardized metric of effect size that allows me to compare two functional ANOVA models?

Let's say I am running two similar ANOVAs. The first ANOVA takes a raw dataset (and computes the differences across groups A, B, and C). The second ANOVA takes a re-scaled version of the same dataset (using the same group variable as factor).

image

Unfortunately, I cannot help to implement any potential solution as my background is in medicine.

Muchísimas gracias por su respuesta y por todo el esfuerzo en el campo del FDA!!!

Alternatives

No response

Additional context

No response

vnmabus commented 2 months ago

Thank you for your question! The functional ANOVA you are referring to was contributed by @DavidGarciaFer a few years back, but I do not remember the details now, and I do not know if he will.

As this seems more of a question about the theoretical aspects of the method, instead of about its implementation in scikit-fda, it is maybe better to contact the paper authors by mail (and if they answer you back, please share that information here, as it might be useful for others!).

I will also mention @jltorrecilla , as it is the scikit-fda contributor that is closer to the paper authors in this case, and maybe knows a little more about this.