readcell is filling cell locations with "1x1 missing"

173 views (last 30 days)
I am using R2109a and it is recommended to use readcell instead of xlsread. Readcell is slower than xlsread and it is returning "1x1 missing" instead of an empty cell when reading an .xlsx file. I don't know what logical I can use to remove these cell locations without brute forcing the problem. Does anyone else know what I can do to fix this?
Capture.PNG
  2 Comments
Stephan Piotrowski
Stephan Piotrowski on 21 Jan 2022
@Josef Shaoul I experienced extremely slow readtable, but found that calling it as readtable(input,'basic',true) greatly improved the performance.

Sign in to comment.

Accepted Answer

Guillaume
Guillaume on 4 Jun 2019
If your data (excluding header) is purely numeric, you'd be better off using readmatrix instead of readcell. If your date is heterogeneous you'd be better off using readtable. The matrix/table will use less memory than a cell array and more crucially make it very easy to remove rows with <missing> entries with rmmissing.
t = rmmissing(readtable(youfile))
With a cell array, rmmissing will only work if the cell array only contain char vectors.
  1 Comment
David Teitge
David Teitge on 6 Jun 2019
Thanks! I had to use readtable, and it does run a lot faster!

Sign in to comment.

More Answers (1)

Jon Martinez
Jon Martinez on 4 Jun 2019
I've had the same problem reading a CSV file and managed to substitute all missing cells with this command:
your_file = readcell(your_path);
your_file(cellfun(@(x) any(ismissing(x)), your_file)) = {''};
The cellfun command will give you a logical array that mantains the structure of the original cell array. Hope it helps.
  2 Comments

Sign in to comment.

Tags

Products


Release

R2019a

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!