Using unstack in order to replicate Pivot Table created in Excel

Hi,

my question is based on the following table:

var1 = {'sec1';'sec2';'sec3';'sec4';'sec5';'sec6';'sec7';'sec8'};
var2 = {'EQ';'EQ';'EQ';'EQ';'FI';'FI';'FI';'FI'};
var3 = {'EQ';'EQ';'CA';'FINALT';'FI';'FI';'CA';'FINALT'};
var4 = {'X1';'X2';'X3';'X4';'X5';'X6';'X7';'X4'};
var5 = [0.50;0.10;0.02;0.10;0.13;0.05;0.05;0.05];

T = table(var1,var2,var3,var4,var5);

Now I would like to replicate the Pivot table that is in the attached xlsx file and that looks like follows

Basically, I am trying to show the values for var5, the columns are grouped by var3 and subgrouped by var4. The rows are aggregated using group sums for var2.

I am trying to use the unstack function as follows:

unstack(T,{'var3','var4'},'var5','GroupingVariables', 'var2','AggregationFunction', @sum)

but I get the following error message:

Error using sum
Invalid data type. First argument must be numeric or logical.

I have tried other combinations, but not being able to replicate the Excel version.

[Mohammad Sami]

The first parameter is the value variable. Since you are using sum, this must be numeric. Your var3 and var4 are non-numeric. I suppose you meant var5. Your var3 and var4 would be the identity variable. Matlab only allows one. So probaly you need to merge var3 and var4.

unstack(T,'var5','var3','GroupingVariables', 'var2','AggregationFunction', @sum)

[SpeedyGonzales]

Thank you Mohammad. Merging var3 and var4 created another error. I think that based on your answer and additional playing around on my side it looks like I can't achieve what I want to do.