How can I add NaNs to a cell array in positions that match the positions of NaNs in a separate matrix?

I have a matrix (A) containing some NaNs. I also have a cell array (B) with no NaNs, containing script in every element.

Where NaNs exist in A, I also want NaNs to appear in B.

Below is my failed attempt to do this. The final line of the script results in an error: "Conversion to cell from double is not possible." Any advice would be much appreciated.

A = rand(2);                          % matrix
A(1,2) = NaN

B = {'fred'   'wilma';                % cell array
     'barney' 'betty'};

nan_locations = find(isnan(A));       % index location of NaNs in A
B(nan_locations) = NaN                % in B insert NaNs in same locations as A

Answer 1

[Andrei Bobrov]

B(isnan(A)) = {NaN}

[Mark]

Bingo. Very many thanks.

Answer 2

[Vishal Rane]

Use

   B{isnan(A)} = NaN

[Vishal Rane]

What you were trying to do was replace an entire cell ( () indexing ) in B by NaN. What you must do is replace the contents ( {} indexing ) of that cell by a NaN.

[Mark]

Vishal. Your solution works perfectly for my example, but for my actual data, when I use your approach, I get the error: "The right hand side of this assignment has too few values to satisfy the left hand side."

My actual A is a double and actual B is cell. All elements in my actual B are strings of very similar length and composition, eg:

>> BACTUAL(1)
ans = 
    'IE0004613620'

[Vishal Rane]

Check if A and B are of identical size.

[Mark]

A <19 x 27 double>
B <19 x 27 cell>

[Vishal Rane]

Above method will give error if there are no NaN values in A.

Post your code if possible.