r-zemblys / irf

Eye-movement event detection using random forest
MIT License
17 stars 11 forks source link

i2mc feature extraction error via MATLAB #4

Open MdShadmanHasan opened 1 year ago

MdShadmanHasan commented 1 year ago

After properly formatting my data according to the internal data format used by irf, when I run the MATLAB script to extract the i2mc feature I get the following error.

nfile =

 1     1

Importing and processing ....\etdata\lookAtPoint_EL_irf\i2mc\tobii_data_i2mc_raw.mat; 1 out of 1 Importing and processing ; Searching for valid interpolation windows Replace interpolation windows with Steffen interpolation 2-Means clustering started for averaged signal Error using filtfilt>getCoeffsAndInitialConditions Data length must be larger than 24 samples.

Error in filtfilt>efiltfilt (line 142) [b2,a2,zi,nfact,L] = getCoeffsAndInitialConditions(b1,a1,Npts);

Error in filtfilt (line 102) y = efiltfilt(b,a,x);

Error in twoClusterWeighting (line 97) ll_d{p+1} = filtfilt(b{p},a{p},ll_d{1});

Error in I2MCfunc (line 159) [data.finalweights,stopped] = twoClusterWeighting(xpos,ypos,missingn,par.downsamples,par.chebyOrder,par.windowtime,par.steptime,par.freq,par.maxerrors);

Error in I2MC_rz (line 49) [fix, finalweights] = I2MCfunc(data,opt);

r-zemblys commented 1 year ago

Error message says "Data length must be larger than 24 samples". Is your data longer than 24 samples?

MdShadmanHasan commented 1 year ago

Yes! The array has 45978 samples!

r-zemblys commented 1 year ago

Well, then the only thing that I can suggest is: try to run training/testing using the data provided in this repository, and compare if your .mat file has the same format as files you get when replicating results