How to calculate a mean of column one across multiple tables?

Hi all,

I have a 21x1 cell with 100x 18 tables (attached).

What I am seeking to do is to calcuate mean of:

row 1, column1, all tables

row 2, column1, all tables

....

row 100, column 1, all tables

Repeat for remaining 17 columns

So ultimately, I will have a table 100x18.

For example: for column 1

P_acc_z = (t{1,1.P_acc_z_meancycle} + t{2,1.P_acc_z_meancycle}, + t{3,1.P_acc_z_meancycle}+ ...., + t{18,1.P_acc_z_meancycle})/18

Can you help please?

Answer 1

[David Hill]

m=sum(arrayfun(@(x)sum(table2array(c{x}(:,1))),1:18))/1800;% c is your cell array, this finds mean of first columns

loop to find means of all columns.

for k=1:18
    m(k)=sum(arrayfun(@(x)sum(table2array(c{x}(:,k))),1:18))/1800;
end

[Tomaszzz]

Thanks@David Hill

Sorry, I did not explain well.

What I am seeking to do is to calcuate mean of:

row 1, column1, all tables

row 2, column1, all tables

....

row 100, column 1, all tables

Repeat for remaining 17 columns

So ultimately, I will have a table 100x18.

Your code gives me a table 1 x18. Sorry for the confusion.

[David Hill]

s=zeros(100,18);
for k=1:18
    s=s+table2array(c{k});
end
s=s/18;

[Tomaszzz]

thanks!

[Tomaszzz]

Hi @David Hill

I have a couple of tables which have only nan values in a single column which results in having nan values for a mean for this particular column. Is there a way to omit nan values?