search

krasvas / EyeMMV

EyeMMV toolbox is a MATLAB toolbox appropriate for post experimental eye movement analysis
GNU General Public License v3.0
26 stars 11 forks source link

NaN stored in the dataset #3

Open JiangTao100 opened 3 years ago

JiangTao100 commented 3 years ago

Thanks for the package. I wanna process my own eye tracking data . However, there will be some 'NaNs' in horizontal and vertical Cartesian coordinates. If I directly run the function 'fixation_detection' ,there will be something wrong in the output .Should I interpolate before I start ?

ryklin commented 3 years ago

If NaN means signal was lost, and it occurs for just a few samples, maybe you can interpolate safely. That’s my intuition.

On Sep 5, 2020, at 7:10 AM, JiangTao100 notifications@github.com wrote:

 Thanks for the package. I wanna process my own eye tracking data . However, there will be some 'NaNs' in horizontal and vertical Cartesian coordinates. If I directly run the function 'fixation_detection' ,there will be something wrong in the output .Should I interpolate before I start ?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe.

krasvas commented 3 years ago

Thanks for using EyeMMV toolbox! EyeMMV’s fixation detection algorithm considers the sequence of the spatiotemporal coordinates of the collected gaze data in order to compute fixation events. Therefore, I suggest you just to remove the NaNs in your raw data before running the corresponding function. You can find more information about the function of the implemented fixation detection algorithm in this research article: Krassanakis, V., Filippakopoulou, V., & Nakos, B. (2014). EyeMMV toolbox: An eye movement post-analysis tool based on a two-step spatial dispersion threshold for fixation identification. Journal of Eye Movement Research, 7(1). DOI: https://doi.org/10.16910/jemr.7.1.1