Environmental Engineering Reference
In-Depth Information
AIC_X2
=
[AIC_TruncNormal_X2 AIC_Lognormal_X2 AIC_TruncGumbel_X2 AIC_
Weibull_X2];
display(AIC_X2);
[AIC_min_X2 Index_X2]
=
min(AIC_X2,[],2);
if Index_X2
=
=
1
display('The best-fit margin for X2 using AIC is TruncNormal
distribution');
elseif Index_X2
=
=
2
display('The best-fit margin for X2 using AIC is Lognormal
distribution');
elseif Index_X2
=
=
3
display('The best-fit margin for X2 using AIC is TruncGumbel
distribution');
elseif Index_X2
=
=
4
display('The best-fit margin for X2 using AIC is Weibull
distribution');
end
%
BIC_X2
=
[BIC_TruncNormal_X2 BIC_Lognormal_X2 BIC_TruncGumbel_X2 BIC_
Weibull_X2];
display(BIC_X2);
[BIC_min_X2 Index_X2]
=
min(BIC_X2,[],2);
if Index_X2
=
=
1
display('The best-fit margin for X2 using BIC is TruncNormal
distribution');
elseif Index_X2
=
=
2
display('The best-fit margin for X2 using BIC is Lognormal
distribution');
elseif Index_X2
=
=
3
display('The best-fit margin for X2 using BIC is TruncGumbel
distribution');
elseif Index_X2
=
=
4
display('The best-fit margin for X2 using BIC is Weibull
distribution');
end
%
% Select best-fit copula among Gaussian,Plackett,Frank and No.16 copulas
% Empirical distributions of data, U
[datasort,dataindex]
=
sort(data);
Ranks_data
=
data;
for m
=
1:cols
Ranks_data(dataindex(:,m),m)
=
1:rows;
end
U
=
Ranks_data/(rows
+
1);
%
% Kendall's tau of data, Kendall_data
Kendall_data
=
corr(data,'type','Kendall');
%
% Estimation of copula parameters
rho_Gaussian
=
sin(pi/2*Kendall_data);
theta_Plackett
=
theta_estimationPlackett_Kendall(Kendall_data(1,2));
theta_Frank
=
copulaparam('Frank',Kendall_data(1,2),'type','Kendall');
theta_16
=
theta_estimation16_Kendall(Kendall_data(1,2));
Copula_parameter
=
[rho_Gaussian(1,2) theta_Plackett theta_Frank
theta_16];
Search WWH ::
Custom Search