The weird thing is that I am actually able to pone s = 0 and x=0 to compute LapI{1,1}(s,x), but I then get the error when I implement it in the for loop
exceeded number of array elements
2 次查看(过去 30 天)
显示 更早的评论
Hello everyone,
could you please help me figure out how to fix this error? It is showing up already for B=1 and C=1
Here's the code:
clear,
% [ THIS IS test_I_analysis ]
%% PARAMETERS
H = [100 ]'; h=1 ; var = 2e5 ;
lambdaA = 1e-6 ; lambdaT = 0.15 ;
lambda = [lambdaA, lambdaA, lambdaT] ;
Sa = [4.88 9.6117 12.0810 27.304]' ; Sb = [0.429 0.1581 0.1139 0.0797]' ;
alpha = [2, 3, 3] ;
alpha= [2.37 3 3]
tau = 0.31623 ; sigma_n2 = 10^-13 ;
p = [1.585, 1.585, 10] ;
eta = [0.9972 0.9943 1]
xi = eta .* p ;
m = [2 1 1] ; % Shape parameters for Nakagami fadings
Ru=600
R= 500 ;
for A=1:2
dd{A,3} = @(d) (xi(3) / xi(A))^(1/alpha(3)) * d.^(alpha(A) / alpha(3)) ;
dd{3,A} = @(d) ( (xi(A) / xi(3))^(1/alpha(A)) * d.^(alpha(3) / alpha(A)) ) .* ( d>=dd{A,3}(H) ) ...
+ H .* ( d<dd{A,3}(H) ) ;
end
dd{2,1} = @(d) (xi(1) / xi(2))^(1/alpha(1)) * d.^(alpha(2) / alpha(1)) ;
dd{1,2} = @(d) ( (xi(2) / xi(1))^(1/alpha(2)) * d.^(alpha(1) / alpha(2)) ) .* ( d>=dd{2,1}(H) ) ...
+ H .* (d<dd{2,1}(H)) ;
d = @(z) sqrt(z.^2 + H^2) ;
for C = 1:3
dd{C,C} = @(z) z ;
for B = 1:3
D{B,C} = @(z) sqrt( dd{B,C}(d(z)).^2 - H^2 ) ;
D{B,3} = @(z) dd{B,3}(d(z)) ;
end
D{3,C} = @(z) sqrt(dd{3,C}(z).^2 - H^2) ;
D{C,C} = @(z) z ;
end
xm = @(Beta) Ru(1)*cos(Beta) - sqrt(R^2 - Ru(1)^2*sin(Beta).^2) ;
BetaX = asin(R/Ru(1)) ;
for u = 1:2
xX{u} = @(Beta) Ru(u)*cos(Beta) + sqrt( R^2 - (Ru(u)*sin(Beta)).^2 ) ;
Betai{u} = @(z) real( (z>=0).*acos( (z.^2 + Ru(u)^2 - R^2) ./ ( (z>0)*2*z*Ru(u) + (z==0)) ) ) ;
end
PLoS = @(z) 1 ./ (1+Sa(1)*exp( -Sb(1)*(180/pi*atan(H./z)-Sa(1)) ) ) ;
integrandK{1} = @(z) z ./ (1+Sa(1)*exp( -Sb(1)*(180/pi*atan(H./z)-Sa(1)) ) ) ;
integrandK{2} = @(z) z .* (1 - 1/(1+Sa(1)*exp( -Sb(1)*(180/pi*atan(H./z)-Sa(1)) ) ) ) ;
for A=1:2
I_{A} = @(s,z) (1 - (m(A)/(m(A) + s .* xi(A) .* (d(z)).^-alpha(A)) ).^m(A) ) .* integrandK{A}(z) ;
for B = 1:3
LapI{B,A} = @(s,x) exp(-2*pi*lambda(A) * integral(@(z) I_{A}(s, z), D{B,A}(x),Ru(u)-R,'ArrayValued',1) ) ...
.* exp(lambda(A)*integral(@(Beta) integral(@(z) I_{A}(s,z), xm(Beta),xX{1}(Beta),'ArrayValued',1), -BetaX,BetaX,'ArrayValued',1 ) ) ;
LapBC{B,A,1} = @(s,x) LapI{B,A}(s, x) .*(Ru(1)-R>D{B,A}(x)) ;
end
end
for u = 1:length(Ru)
r2{u} = @(Beta,z) Ru(u)^2 + z.^2 - 2*Ru(u)*z.*cos(Beta) ;
I_T{u} = @(Beta,s,z) exp(-r2{u}(Beta,z)./(2*var)) .* (1-(m(3)/(m(3)+s.*xi(3).*z.^-alpha(3))).^m(3)) .* z ;
for B=1:3
LapBC{B,3,u} = @(s,x) exp(-lambda(3)/sqrt(2*pi*var) * integral(@(z) integral(@(Beta) I_T{u}(Beta,s,z), 0,2*pi,'ArrayValued',1), D{B,3}(x), Ru(u)+R,'ArrayValued',1) ) ;
LapB{B,u} = @(s,x) LapBC{B,1,u}(s,x) .* LapBC{B,2,u}(s,x) .* LapBC{B,3,u}(s,x) ;
end
end
% [THIS IS test_I]
u=1;
DsI = [0 0 0] ;
xLapI = zeros(3);
LapBC_ = {1 1 1; 1 1 1; 1 1 1}
for B=1:2
for C = 1:2
LapBC_{B,C} = LapBC{B,C,u}(tau/xi(B)*DsI(B)^alpha(B), xLapI(B,C)) ;
end
end
回答(1 个)
Stephen23
2020-10-22
Ru is scalar, but in multiple locations you try to access Ru(2), e.g.:
for u = 1:2
xX{u} = @(Beta) Ru(u)*cos(Beta) + sqrt( R^2 - (Ru(u)*sin(Beta)).^2 ) ;
% ^^^^^ ^^^^^
and
LapI{B,A} = @(s,x) exp(-2*pi*lambda(A) * integral(@(z) I_{A}(s, z), D{B,A}(x),Ru(u)-R,'ArrayValued',1) ) ...
% ^^^^^
2 个评论
Stephen23
2020-10-22
Funnily enough I used Octave, because it gives an error message which names the exact variable related to the error:
̀error: Ru(2): out of bound 1
error: called from
temp0>@<anonymous> at line 56 column 23
temp0>@<anonymous> at line 58 column 39
temp0 at line 78 column 17
The MATLAB Editor is great for most debugging, but for cases like this just the line number is not enough.
另请参阅
类别
在 Help Center 和 File Exchange 中查找有关 Point Cloud Processing 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!