Convert multiple Table Variables from Cell to Double

I have a table with many variables of mixed types (some strings, some numbers). The first X variables should be double/numeric, but they are cell arrays and in each cell the number is bounded by single quotes ({'2.496' }, {'6.675' } ,{'9.644' }, etc). I can convert the X columns to double one at at time.

T.bin1 = str2double(T.bin1);

T.bin2 = str2double(T.bin2);

...

T.binX = str2double(T.binX);

I checked the output: T.bin1 is a 167x1 cell array. str2double turns it into a 167x1 double. This single command methods works fine. But, I'm having trouble generalizing the code to a situation where I saved table names: HoldNames = T.Properties.VariableNames([1:X]);

This will work, but it's one at a time:

T.bin1 = str2double(table2cell(T(:,HoldNames(:,1)))) %a 167x1 double assigned into a 167 x1 Table

T.bin2 = str2double(table2cell(T(:,HoldNames(:,2))))

...

T.binX = str2double(table2cell(T(:,HoldNames(:,X))))

However, these do not work:

%a 167xX double assigned into a 167xX Table

T(:,HoldNames) = str2double(table2cell(T(:,HoldNames))) %Right hand side of an assignment into a table must be another table or a cell array.

%a 167xX double assigned into a 167xX cell array

T{:,HoldNames} = str2double(table2cell(T(:,HoldNames))) %Conversion to cell from double is not possible.

Answer 1

[Cris LaPierre]

I was able to get the following dummy example to work.

% Create table of cells
var = num2cell(rand(4,3));
var = cellfun(@num2str,var,'UniformOutput',false);
T = splitvars(table(var))

T = 4x3 table
       var_1           var_2          var_3    
    ____________    ___________    ____________

    {'0.15345' }    {'0.44518'}    {'0.93558' }
    {'0.095759'}    {'0.86483'}    {'0.55891' }
    {'0.081027'}    {'0.92421'}    {'0.068477'}
    {'0.95817' }    {'0.35516'}    {'0.076722'}

% Convert Var_2 and Var_3 to double (still cells)
HoldNames=["var_2","var_3"];
T{:,HoldNames} = table2cell(varfun(@str2double,T,'InputVariables',HoldNames))

T = 4x3 table
       var_1          var_2         var_3   
    ____________    __________    __________

    {'0.15345' }    {[0.4452]}    {[0.9356]}
    {'0.095759'}    {[0.8648]}    {[0.5589]}
    {'0.081027'}    {[0.9242]}    {[0.0685]}
    {'0.95817' }    {[0.3552]}    {[0.0767]}

[Cris LaPierre]

I don't like that they are still cells. This approach isn't pretty, but it's the first way I found to change that.

% Create table of cells
var = num2cell(rand(4,3));
var = cellfun(@num2str,var,'UniformOutput',false);
T = splitvars(table(var))

T = 4x3 table
       var_1          var_2          var_3   
    ___________    ___________    ___________

    {'0.68742'}    {'0.25061'}    {'0.41257'}
    {'0.60281'}    {'0.30769'}    {'0.98267'}
    {'0.23292'}    {'0.70099'}    {'0.93851'}
    {'0.26057'}    {'0.43727'}    {'0.65197'}

% Convert Var_2 and Var_3 to double (as arrays)
HoldNames=["var_2","var_3"];
T_temp = varfun(@str2double,(varfun(@string,T(:,HoldNames))));
T(:,HoldNames)=[];
T(:,HoldNames)=T_temp

T = 4x3 table
       var_1        var_2      var_3 
    ___________    _______    _______

    {'0.68742'}    0.25061    0.41257
    {'0.60281'}    0.30769    0.98267
    {'0.23292'}    0.70099    0.93851
    {'0.26057'}    0.43727    0.65197