In fi(var1,var2,...varN) if var1 is not a constant then

Hello There,

I want your help please to convert a double array to fixed point and keep the FractionLength of each variable in array separate

I'm trying to convert a variable array (wi) from double to fixed point. The variables of the array have a different FractionLength fraction as showing below, and when I do "wi=fi (wi,1)"; it takes a round FractionLength for the entire array which affect the values of each variable.

>> fi(wi(1,1))
ans = 

      0.3956

            DataTypeMode: Fixed-point: binary point scaling
              Signedness: Signed
              WordLength: 16
          FractionLength: 16

>> fi(wi(1,3))
ans = 

     -0.1145

            DataTypeMode: Fixed-point: binary point scaling
              Signedness: Signed
              WordLength: 16
          FractionLength: 18

but if I do a fixed point for the entire array (fi(wi))

>> fi(wi)

            DataTypeMode: Fixed-point: binary point scaling
              Signedness: Signed
              WordLength: 16
          FractionLength: 13

I tried to do the conversation of fixed point without indicating the FractionLength as below, but I got this error message.

wi=fi (wi,1,16)

Error "In fi(var1,var2,...varN) if var1 is not a constant then var2 to varN must be or specify a complete numerictype."

Can you please help me to do this on the right way?

Thank you in advance

Answer 1

[Jeevan Joishi]

The fixed-point properties WordLength and FractionLength are part of a fixed-point number's datatype. Fixed-point numbers with different FractionLength/WordLength are considered different data types.

In MATLAB, an array can be used to store multiple elements of the same datatype. Therefore an array cannot be used to store fixed-point numbers of different WordLength/FractionLength because their data types are different.

Alternatively, a cell array can be used to store fixed-point number with different WordLength/FractionLength.

For instance,

>> wi = {16 2 3 13; 5 11 10 8; 9 7 6 12; 4 14 15 1};

>> fiOut = cellfun(@fi,wi,'UniformOutput',false);

It is not possible to apply 'fi' directly on the entire cell array. Thus you can use 'cellfun' to apply the function to each element of the cell array 'wi'.

[Life is Wonderful]

wi = {16 2 3 13; 5 11 10 8; 9 7 6 12; 4 14 15 1};
fiOut = cellfun(@fi,wi,'UniformOutput','false');

Error using cellfun
Input #4 expected to be a logical value for the parameter 'UniformOutput'.

fiOut

[Stephen23]

@Jogger: the code given in the answer was clearly incorrect and does not work. I have corrected it.

As the error message states, rather than using a character vector 'false' you should use a logical value:

wi = {16 2 3 13; 5 11 10 8; 9 7 6 12; 4 14 15 1};
fiOut = cellfun(@fi,wi,'UniformOutput',false)

fiOut = 4×4 cell array
    {1×1 embedded.fi}    {1×1 embedded.fi}    {1×1 embedded.fi}    {1×1 embedded.fi}
    {1×1 embedded.fi}    {1×1 embedded.fi}    {1×1 embedded.fi}    {1×1 embedded.fi}
    {1×1 embedded.fi}    {1×1 embedded.fi}    {1×1 embedded.fi}    {1×1 embedded.fi}
    {1×1 embedded.fi}    {1×1 embedded.fi}    {1×1 embedded.fi}    {1×1 embedded.fi}