Readmatrix is changing the "size" of the data I am trying to retrieve from an Excel file?

Sort of a beginner in MatLab here.

I have a very large Excel file that I am trying to read, but I'm only concerned with two columns right now. Both columns have the exact same row range, and I should end up with a nice 6000x1 matrix. However, my second matrix (B1f) is not coming out with this size.

Please note that I'm trying to use the fillmissing command to replace all blank cells with a zero (there are a LOT of empty cells) to aid in not dealing with NaN and whatnot.

Here's the code I've got so far:

[file,filepath] = uigetfile ('*.xlsx');
file = fullfile(filepath,file);

noise = readmatrix(file,'Sheet','continuous-waveforms','Range','B2:B6001');
minValue = min(noise(:));
truezero = noise - minValue;

B1peak = readmatrix(file,'Sheet','continuous-waveforms','Range','CY2:CY6001');
B1f = fillmissing(B1peak,'constant',0); 
B1ffix = B1f - truezero; %This is where the different sized matrices occur.
B1pos = B1ffix(B1ffix > 0); 
MeanB1 = mean(B1pos); 

I don't get any error code, just that "Matrix dimensions must agree", which, duh. My variables in the workspace window confirms the significant difference in matrix sizes.

I've been fighting with this for a while. The same thing was happening with xlsread but I tried switching to readmatrix because it seemed like a little bit better for this.

[Walter Roberson]

I do not seem to be able to replicate that on R2021a on Mac.

One thing I would ask is whether B1peak is coming out smaller than expected -- compared to the possibility that fillmissing() could be the culprit (hypothetically)

Which MATLAB release are you using, on which OS? And if you are using Windows, do you have Excel installed?

[Madeline Stalder]

Yes, B1peak is coming out smaller than it should (5447x1). Also, the size doesn't change after my fillmissing() goes through.

I am using R2020a on Windows 10 and Excel is installed. I can attach the Excel file I'm trying to read right now to see if yours still is good on your end.

I'm wondering if I should instead try to just get matlab to pre-fill in my zeros for me? Or maybe I should just try to download the newest version of MatLab.

[Walter Roberson]

You said "a very large" file; if it is more than 5 megabytes you might need to put it into Google Drive or DropBox and post the link.

I do not have Excel installed in Windows, but I could test anyhow.

Answer 1

[Sulaymon Eshkabilov]

Hi,

This is where you have logical indexing that is removing all data points which are less than 0.

B1pos = B1ffix(B1ffix > 0); 
MeanB1 = mean(B1pos); 

[Madeline Stalder]

But since this line comes after the matrix subtraction, it shouldn't affect it right?