CST Active Element Pattern

Hi guys,
I am trying to model an array system and considering the active element pattern approch to include the mutual coupling effect. For this I tried two different approches.
1) I ran a small array simulation with all ports setting and then determined the Z matrix using the CST. I took a stand alone element pattern and multiply it with the general array factor. Then I chose a set of weights e.g p=[1 1 1 1] and multiply it with the Z matrix. Then these corrupted weights are multiplied to the whole array factor;
A=stand alone element pattern
p=[1 1 1 1]
Z=Impedence matrix
I=Z*p corrupted with coupling
TF= sum(r=1:4) I_r *A(theta)*AF(theta) where AF=array factor; TF = total array factor
Now using Matlab I got the pattern with all the coupling effects. But when I apply the same weights p=[1 1 1 1] to CST model it gives a difefrent pattern ... What did I miss here?
2) Secondly I am using individual element pattern in the presence of others. so now the coupling effect is on the individual element pattern. Again I use Matlab, determined the pattern;
TF=p*sum(r=1:4)A_r*AF
still the results are not the same as on CST.
Any one has any idea about it?
Thanks in advance
Cheers

Hi Asim Ali Khan,
Did you solve your problem.
I also am studying about phased array of patch antennas including mutual coupling.
So, if you solved your problem, plz share.
Thank you so much!
Rgs