Benedikt Ehinger
Benedikt Ehinger
Draft Pullrequest that contains the edits we are doing to get the pipeline working with newest numpy/python + the trackpixx eyetracker
here we flip gy for eyelink: https://github.com/behinger/etcomp/blob/f30389da49c3416c7a723d44951d197d6e89d40e/code/functions/et_import.py#L314 but we forgot to flip the velocity which is calculated by the eyelink directly. Thus if you calculate the velocity yourself, it will...
in another study we need more robust eye detection than the one PL offers. https://www.sciencedirect.com/science/article/pii/S0165027019301578?via%3Dihub
from papr: ```python import msgpack gaze_serialized = [msgpack.packb(gaze, use_bin_type=True) for gaze in gaze_list] ``` In order to addapt to their new style
https://www.nature.com/articles/s41598-017-17983-x.pdf https://gitlab.com/nslr/nslr-hmm/blob/master/nslr_hmm.py
Widmann et al 2014 recomends not to use filtfilt, because cut-off frequency becomes dificult to estimate etc. I also wouldn't use Butterworth but FIRWindow with a Hamming window; mirroring the...
I have a reject data function in Unfold.jl [here](https://github.com/unfoldtoolbox/Unfold.jl/blob/9612d63019c1bb5ce0052d70fb706fb80fff895c/src/utilities.jl#L106) that simply puts rejected times to `Missing` instead of outright removing it. There is a second helper function to actually remove...
@juliaregistrator register Unfold 0.4.0 Release notes **New feature:** Fit multiple mass-univariate models at the same time. Similar to how you can fit multiple events/basisfunction in an UnfoldLinearModelContinuousTime, you can now...
- improved designmatrix due to interpolate=false, speedup 3x - up to 10x faster designmat for FIR designs - fix unittests? - double speedup due to unnecessary copying in some cases...
I saw in simple testing a ~12% increase in performance just by scaling ```julia function scale_ls!(A) s = ones(size(A, 2)) for j = 1 : size(A, 2) i = A.colptr[j]...