How to normalize size of smaller arrays to match bigger array for analysis, and remove NaNs?

So I have an excel file (attached) that has different trials, all having different column sizes. I want to plot the average of these trials by agent (dex, propofol and alfax), so I should get three different lines for three different averages on the same plot. On the plot I want to display time on the x-axis (excel file: test_analyzingtimeHR.xlsx) and on the y-axis, i want to show the average HR (test_analyzingHR.xlsx).

However, this is being difficult because all these trials have different column sizes (WITHOUT NaNs). The largest column size without any NaNs is 1779. For this reason, I want to normalize all the trials so that all the trials have the same column size (1779). I also don't want any NaNs to show up in my matrices, as they're disrupting my calculations for averaging. Can anyone tell me how I could do this best?

I've attached my Matlab code in the attachments.

Answer 1

[Matt J]

Removing the NaNs is the wrong approach. Just use 'omitnan' flags where appropriate. For example,

A =

    0.5637    0.8022    0.4707
    0.9468    0.8322    0.5755
    0.3354    0.5464    0.9030
       NaN    0.4492       NaN
       NaN    0.4970    0.4532
       
>> mean(A,'omitnan')

ans =

    0.6153    0.6254    0.6006