filter option in signal.clean is not exposed to nilearn.maskers.NiftiMasker and potentially other masker objects

[htwangtw]

However, I don't think a filter option is provided for nilearn.maskers.NiftiMasker so seems it's using default aka butterworth

Originally posted by @DasDominus in #3434 (comment)

[emdupre]

I think this would need to be modified here.

But since low_pass isn't supported for cosine filtering, we might want to also add a check for sensible args ?

[TakshDhabalia]

Hey , Would like to know about this further and know more about this and if it is assigned or not. I am a beginner here so could you explain a bit more on this?

Thanks!

[bthirion]

IIUC, you can try and add the parameter, and then check possible inconsistencies that may arise.
For that you need to understand the hierarchy of class and functional calls. Maybe @ymzayek you could have a quick meeting together to review the question ?

[ymzayek]

@TakshDhabalia thanks for your interest in working on this! You can start as suggested by @emdupre on where to modify the code and additionally you need to include the argument in the maskers' classes __init__ method. You can start with just one masker for simplicity. Also what could help you get started is to take as an example another parameter of signal.clean that is already exposed in e.g. nilearn.maskers.NiftiMasker such as low_pass or high_pass and copy that implementation for filter. Does it make sense?

[TakshDhabalia]

yep , I agree with that but I need a little more time on reading the documentation as I am struggling a bit on this project as a whole , I will get started on it right away but if anyone wants to help , it would be very much appreciated

[ymzayek]

Hi @TakshDhabalia you can certainly take your time with this and we are here to help you along the way. If you have any specific questions whether it has to do with this issue or more generally about contributing to nilearn or how we structure our code, etc. we are happy to give you tips and point you to the right resources. You can already find a lot of information here to help you get started: https://nilearn.github.io/stable/development.html. And once you open a PR we can help you along the way with the code. You can also directly talk to us during our office hours meeting (info in the link above but check https://twitter.com/nilearn for the most up-to-date meeting information). We are very happy that you're interested in contributing to nilearn :)