Find unique column values in NxN cell array

I have a 1000x50 cell array with each column contining multiple duplicate entries.

simplified example:

names = {'red','blue','green';'red','orange','yellow';'orange','orange','red'};

{'red' } {'blue' } {'green' }

{'red' } {'orange'} {'yellow'}

{'orange'} {'orange'} {'red' }

I would to create a simplified array wich contains only the unique values from each of the indiviual columns:

result:

{'red' } {'blue' } {'green' }

{'orange'} {'orange'} {'yellow'}

{'red' }

[tybo1mos]

My original data is actually in table form. So my prference would be to work from the original table data, and rutun a table with only unique values.

Starting:

color_1     color_2     color_3 
________    ________    ________

'red'       'blue'      'green' 
'red'       'orange'    'yellow'
'orange'    'orange'    'red'  

Reult

color_1     color_2     color_3 
_______    ________    ________

'red'       'blue'      'green' 
'orange'    'orange'    'yellow'
                        'red'  

[Paul]

Is that result feasible? I thought each table variable has to have the same number of rows?

Answer 1

[Paul]

Because the resulting variables could all have a different number of elements, I'm converting each variable into a scalar cell that contains the desired data

t = table({'red';'red';'orange'},{'blue';'orange';'orange'},{'green';'yellow';'red'})

t = 3×3 table
       Var1          Var2          Var3   
    __________    __________    __________

    {'red'   }    {'blue'  }    {'green' }
    {'red'   }    {'orange'}    {'yellow'}
    {'orange'}    {'orange'}    {'red'   }

t = varfun(@(x){unique(x,'stable')},t)

t = 1×3 table
     Fun_Var1      Fun_Var2      Fun_Var3 
    __________    __________    __________

    {2×1 cell}    {2×1 cell}    {3×1 cell}

t.(1){:}

ans = 2×1 cell array
    {'red'   }
    {'orange'}

t.(2){:}

ans = 2×1 cell array
    {'blue'  }
    {'orange'}

t.(3){:}

ans = 3×1 cell array
    {'green' }
    {'yellow'}
    {'red'   }