How to preallocate the arrays(mat​rix1,matri​x2,matrix3​) and variables(​trapdensit​ycount,bia​s,beta1,be​ta2) in the following question

%clears variables, closes windows and clears screen clear all; close all; clc;

%defining variables T= 310; % absolute temperature in Kelvin Dn=1.0; Dp=0.3; Nd=1e18; Na=1e18; epsilonn=11.8; epsilonp=11.8; ni=1.5e10;

sigma_n=1e-5; sigma_p=1e-6;

% T1=0;T2=0;T3=0;T4=0;T5=0; Vth=0.026*T/300 ; window_width=1E-6;

alpha1 = 1.0;

%creating vectors len_start=0; len_end = 99; len_num_points =100;

vec_of_lengths = linspace(len_start,len_end, len_num_points);

rad_start=1; rad_end =100 ; rad_num_points = 100;

EField_num_points =100 ;

for count = 1:length(vec_of_lengths)

vec_of_rad = linspace(rad_start,rad_end, rad_num_points); vec__of_aspect_ratios = vec_of_lengths./vec_of_rad;

V_built_in=0.7;% built-in voltage in volts

volt_start = 0; volt_end = 0.95*V_built_in; volt_num_points = 100;

vec_of_volts = linspace(volt_start, volt_end, volt_num_points); length_vec_of_volts = length(vec_of_volts);

matrix1=[];

for count2= 1:17 matrix1(:,:,count2) = zeros(length_vec_of_volts, rad_num_points); end

for count2=1:4 matrix2(:,:,:,count2) = zeros(EField_num_points, rad_num_points, length_vec_of_volts); end

for count2=1:2 matrix3(:,count2) = zeros(1, rad_num_points); end

for count2=1:length(vec_of_rad)

for count3=1:length(vec_of_volts)

Ln=vec_of_rad(count3);% Ln is the diffusion length of electrons and is set equal to the radius.

Nr=Dn/(Ln*Ln*sigma_n*Vth);

taun0 = 1/(sigma_n*Nr*Vth);

taup0 = 1/(sigma_p*Nr*Vth);

taun=taun0;

taup=taup0;

trapdensitycount3(count3,count2)=taun0;

Lp=sqrt(taup*Dp);

bias(count3,count2) = vec_of_volts(count2); % assigning bias a value from voltage vector

%call depletion region function [x1,x2,x3,x4,V_built_in] = depletionregion(Nd, Na,vec_of_rad(count2),epsilonp, epsilonn, ni,bias(count3,count2), window_width)

if(bias(count3,count2) > V_built_in ) %V_built_in(count3,count2))

break;

end %end if

%add x1 and x2 d1=x1 + x2;

%add x3 and x4 d2=x3 + x4;

% define dimensionless variables

 beta1(count3,count2) = vec_of_rad(count2)/Lp;

 beta2(count3,count2) = (vec_of_rad(count2) - x1)/Lp;  % x1(count3,count2))/Lp; this commented part has to be used after 

 beta3 = alpha1*vec_of_lengths(count2);

end

end

end

Answer 1

[Geoff Hayes]

Look at how you are updating Matrix1

 matrix1=[];

 for count2= 1:17 
    matrix1(:,:,count2) = zeros(length_vec_of_volts, rad_num_points); 
 end

The third dimension is based on count2, so must be 17. Within each matrix1(:,:,count2) we insert a zeros matrix with length_vec_of_volt rows and rad_num_points columns. So we can initialize matrix1 as

 matrix1 = zeros(length_vec_of_volts, rad_num_points, 17);

The same can be done for matrix2, matrix3, and the other variables. However, given that your code only initializes the matrix1, matrix2, and matrix3 and doesn't make use of them, it isn't clear why you need to do this.

[Sai Priya]

Thank You!

But how to preallocate the variables?

[Geoff Hayes]

The same way - the other variables, bias1, bias2, etc. appear to be arrays too.