Delete values in a table but keep the cells

I have a 1846x72 table that contains strings (I assume they are strings since the value is in '') and numbers. The first two columns are either NaN or a number. I want to go through each row and, if the second column in that row is not NaN, make the following columns all NaN (or empty) for only that row.

This is the code I a have:

for r = 1:nRows

if isnumeric(df{r,2})

for c = 3:nCols

df{r,c} = NaN;

end

end

end

This is the error I am receiving:

Conversion to cell from double is not possible.

I can not include the dataset as it pertains to my work and I am not allowed.

[Daniel Garratt]

I do not actually want to make every one of the following columns NaN but I feel if I get this sorted then I can figure it out from there.

[Simon Allosserie]

Use df{r,c} = [], that should work

[Daniel Garratt]

That returns the error:

"The value on the right-hand side of the assignment has the wrong width. The assignment requires a value whose width is 1."

Answer 1

[Jorg Woehl]

Try

df{r,c} = {''}

instead. You should be able to avoid the inner loop entirely by using

df{r,3:nCols} = {''}

Also, to check if the second column in row r is not NaN, use

if ~isnan(df{r,2})
    ...
end

isnumeric(NaN) will return logical true, as NaNs belong to the floating point class.

[Daniel Garratt]

I like how much more compact that it. Unfortunately, that returns yet another error:

"Conversion to double from cell is not possible."

I am now trying to convert the whole table to a cell array but I'm not sure if that'll do the trick or why not every column with numbers as entries is being treated as a double.

[Daniel Garratt]

It took a little changing but this is what I have working.

% import data

df = readtable('path to file');

cellarray = table2cell(df);

[nRows, nCols] = size(df);

% go through each row and empty all remaining cells of each row after the

% second column if the second column is not equal to NAN

for r = 1:nRows

if ~isnan(cellarray{r,2})

for c = 3:nCols

cellarray{r,c} = {''};

end

end

end

Thanks for your help.

[Jorg Woehl]

I'm glad you were able to make it work! Not sure why you've still received a type conversion error when working with the table... probably something specific to the imported table.

If you can provide a small table with (made-up) sample data, we might be able to figure out what is going on. But the main thing is that your code is running.