cellfun error when using

Hi everyone, can you help me for my code? this is my code:

clear all
samples = [    88    93    82    39    58    55    90    68     4    78    85    95    24    57     8    68
  66    88    45    16    24    67    32    66    97    78    85    58    73    16    37    18
   6    32    59    44    81    22    31    94    24    61    68    91    88    62    79    66
 100    93    13    83    39    37    29    28    22    80    59    79    22    85    92    97
   8    50    21     6    14    84    93    34    20     4    51    60   100    72    55    13
  39    10     8    49    13    52    69    47    40    73    42    45    83     9    58    49
  89    20    82    84    37    74    68    56    54    93    99    36    44    90    26    50
  37    15    73    73    11    37    33    19    97    25    21    75    56    85    68    83
  98    12    62    61    59    78    39     9    78    21    53    25    74    21    73    20
  52    84    87    58    33     5    38    99    88    21    16    15    73    42     6    23
  67     3    55    10    20    44    30    11    53    85    41    46    23    71    46    79
  72    79    16    38    86    88    19    69    43    64    21    30    74    57     8    32
 100    75    64    61    86     7    93    91    43    44    34    49    81    30    95    55
  53    14    84    89    62     8    24    63   100    48    63    54    80    16    81    88
   8    61    73    96    54     5    83    96    59    90    72    75    98    40    71    76
 100     6    17    21    32    65     6    68    99    29    79    18    20    70    49    85];
threshold = 90;

for r=1:16
      for c=1:16
          CellActivate(r,c) = {samples(r,c)};
          CellConfig.Boosts{r,c} = rand(2,2);
          CellConfig.value{r,c} = 2*ones(2,2);
      end
end
E = cellfun( @(B,C) B(C>threshold)+ones(2,2) , CellConfig.Boosts, CellActivate,'UniformOutput',false);

my point is E will return value of B( when C > threshold) plus ones(2,2).

Thank for your help.

Answer 1

[Walter Roberson]

Each of your CellActivate entries is a scalar. Your threshold is a scalar. For any one entry in CellActivate, comparing it to a scalar is going to get you a scalar true or false.

Your CellConfig.Boosts entries are 2 x 2. Indexing a 2 x 2 at a scalar logical is going to get you either emptiness or a scalar double.

This value that is either emptiness or a scalar double is added to a 2 x 2 array. It is valid to add a scalar to a 2 x 2 array, getting a 2 x 2 array. But it is not valid to add emptiness to a 2 x 2 array: you will get an error.

"my point is E will return value of B( when C > threshold) plus ones(2,2)."

You might mean one of two things there:

1. That you want as output as many cells as in the original, that are the same size as the original, but in the cell entries corresponding to the cases where the threshold is reached, you want to add ones(2,2) to those entries; or
2. That you want to use the threshold to select a subset of the B cells, and you want to add ones(2,2) to each member of the subset.

The second of those is easier to implement:

E = cellfun(@(B) B+ones(2,2), cellConfig.Boosts( cellfun(@(C) C>threshold, CellActivate) ), 'uniform', 0);

[Walter Roberson]

For the first of those,

mask = cellfun(@(C) C>threshold, CellActivate);
E = cellConfig.Boosts;
E(mask) = cellfun(@(B) B+ones(2,2), E(mask), 'uniform', 0);

Answer 2

[Tiki Tiki]

Thank.

I run your code. it is good. but because I plan to use vectorizing, so my output E should be same size with CellConfig.

so Can I use vectorizing here?