Some refactoring

[Gilles86]

Hey @tknapen /maybe @jwdegee ,

I am currently analyzing pupil data from Zurich and therefore refactoring the hedfpy code a little bit. First thing to do is make it load at all in Python 3. I also don't like to us CapitalLetterInCamelCase in filenames, so I changed that.

Next step is a function that allows to analyze data that was not made by exptools/Knapenlab etc.

[tknapen]

Hey Gilles, I know that the cc’d uva people are also looking at making a new version of hedfpy. Perhaps you’d like to contribute to a more comprehensive overhaul at some point? Cheers, T

…

Sent from my iPad

On 7 Jan 2020, at 21:23, Gilles de Hollander ***@***.***> wrote:  Hey @tknapen /maybe @jwdegee , I am currently analyzing pupil data from Zurich and therefore refactoring the hedfpy code a little bit. First thing to do is make it load at all in Python 3. I also don't like to us CapitalLetterInCamelCase in filenames, so I changed that. Next step is a function that allows to analyze data that was not made by exptools/Knapenlab etc. You can view, comment on, or merge this pull request online at: #6 Commit Summary Upper case in filenames is no-go File Changes M hedfpy/Operator.py (5) M hedfpy/__init__.py (6) R hedfpy/commandline.py (6) R hedfpy/edf.py (6) R hedfpy/eyesignal.py (2) R hedfpy/hdf.py (11) Patch Links: https://github.com/tknapen/hedfpy/pull/6.patch https://github.com/tknapen/hedfpy/pull/6.diff — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

[Gilles86]

Ah!

I'm happy to contribute. I'm positively surprised how quickly I got it to work now with the data from Zurich: I think all the core-functionality is there, but the API is just a bit specific for how you do things and hard to understand coming from outside. Nothing that can't be fixed in a few days.

UvA-people: where are you currently with this?

[Gilles86]

There is also a dependency on the oldschool fir-package that has to be addressed.

[lukassnoek]

Ah, great work Gilles. If your PR are regularly merged back, we can also start working from this repo (with a separate fork). How about the HDF output? Should we keep that or change it to a more BIDS-like output (tsv.gz + metadata)?

[Gilles86]

If your PR are regularly merged back, we can also start working from this repo (with a separate fork).
Do you have a different fork?

How` about the HDF output? Should we keep that or change it to a more BIDS-like output (tsv.gz + metadata)?
I'm not sure. I don't particularly like the HDF format to work with in analyses. However, it is kinda nice to have it as a data backbone of the preprocessing class.

Maybe we can have a Skype call about this and make some decisions?

[lukassnoek]

Sure. Let's continue this on Slack.