How to convert cell 2 double?

>> U=SI
U =
  1×2 cell array
    [1×2 double]    [1×2 double]
>> U{1}
ans =
     1     2
>> U{2}
ans =
     2     3
>> sitem=[]
sitem =
     []
>> class(sitem)
ans =
    'double'

I want to copy each cell of U in sitem which sitem remain double class

[dpb]

sitem=cell2mat(U);

Why the need for U at all, though? Just use SI since (apparently) already have it.

[Mira le]

I need U because its values are changing during the execution

[Mira le]

an error appear :

Error using cell2mat (line 52)

CELL2MAT does not support cell arrays containing cell arrays or objects.

[dpb]

Well, so would SI then, wouldn't it?

Did cell2mat solve your problem, though?

[dpb]

You've apparently gotten cell arrays embedded in cell arrays...we'll then have to see just what you actually have to know how many levels of dereferencing it'll take. Or, better yet, remove one or more unecessary levels.

As starters, though, what does

whos SI U

return? And, for good measure, just

SI

at the command line?

In addition, show us the code that creates SI.

[Mira le]

SI is cell array

SI

SI =

1×2 cell array

[1×2 double] [1×2 double]

sitem should be doule , and each time sitem has one value of SI or all value of SI

like that;

sitem= unique([U{:}])

Answer 1

[dpb]

If SI is as you say...

>> SI={rand(1,2) rand(1,2)}
SI =
  1×2 cell array
    {1×2 double}    {1×2 double}
>> sitem=cell2mat(SI)
sitem =
    0.7150    0.9037    0.8909    0.3342
>> 

then cell2mat does the dirty of the original question of returning array of all elements of SI.

You didn't say nothing about no unique() being anywhere in the picture before, altho it's unclear it has any bearing on the question asked.

The "{:}" idiom/syntax returns a comma-separated list of the elements in the cell array, hence requiring the [] to catenate together as argument. Either is OK, cell2mat is probably somewhat more efficient than [].

NB: Either will run into problems if the sizes of the cell arrays are disparate in dimensions as both are a catenation operation.

In the end, it isn't fully clear what the actual question really was/is...