How do I remove NaN values from a cell array?

I'm editing some code I was given to read in spreadsheet data collected in the field into Matlab. Some of the cells in these spreadsheets are merged, and so when I read those in, I get cells that look like ['number' 'NaN' 'number' 'NaN' ...]. I want to convert the cells to matrices, but cell2mat doesn't work when the cells contain data other than numbers. I'm trying to use "rmmissing", but get the following error:

I suppose the obvious solution would be to unmerge the cells in the spreadsheets, but there's a lot of them, and I want to see if there's a quicker way I can do this in Matlab once that data is imported.

Answer 1

[Matt J]

The presence of NaNs isn't the problem. The problem is that the cell contents are character vectors of different lengths. You can convert them like so,

>>   str2double({'1','2';'NaN','40'}) 

ans =

     1     2
   NaN    40