How to makedisk for a three parameter weibull distribution

Everytime I run my code with below lines, it displays "Error using makedist (line 101). Invalid parameter name: t0"

t0  = 0.5;
b   = 1:5;
T   = 1:5;
dist = makedist('weibull','a',T,'b',b,'c',t0)
[h,p,adstat,cv] = adtest(data,'Distribution',dist)

How can I integrate a three parameter weibull distribution into "makedist"?

Answer 1

[Jeff Miller]

MATLAB's Weibull distribution only allows 2 parameters, so you must handle the offset separately. Try this:

t0  = 0.5; % 
b   = 1.5; % you are using vector 
T   = 1.5; % same here too 
dist = makedist('weibull','a',T,'b',b)
dataAdjusted = data - t0;
[h,p,adstat,cv] = adtest(dataAdjusted,'Distribution',dist)

[Mustafa Vural]

I am generating random numbers of a three parameter weibull distribution. I want to run the Anderson Darling test on my generated results. so "data" are my generated result. How can I do this in this case?

clear all;
n   = 100;
t0  = 0.5;
b = 1:5;
T = 1:5;
for v_T= T
    for v_b= b
        data(:,v_b,v_T) = wblrnd(v_b,v_T, [n,1]) + t0;
    end
end

[VBBV]

%if true
  % code
%end
clear all;
n   = 100;
t0  = 0.5;
b = 1:5;
T = 1:5;
for v_T= 1:length(T)
  for v_b= 1:length(b)

Use the for loop as shown above or as below

%if true
  % code
%end
for v_T = [T]
 for v_b = [b]

[Mustafa Vural]

So like this? Sry I am very noob in matlab.

clear all;
n   = 100;
t0  = 0.5;
b = 1:5;
T = 1:5;
for v_T= 1:length(T)
  for v_b= 1:length(b)
        data(:,v_b,v_T) = wblrnd(v_b,v_T, [n,1]) + t0;
  end
end
 for v_T= 1:length(T)
  for v_b= 1:length(b)
dist = makedist('weibull','a',v_T,'b',v_b)
dataAdjusted = data - t0;
[h,p,adstat,cv] = adtest(dataAdjusted,'Distribution',dist)
  end 
 end

Answer 2

[VBBV]

you are using vector of values for 2 parameters and scalar value for other parameter ... try this

%if true
  % code
%end
t0  = 0.5; % 
b   = 1.5; % you are using vector 
T   = 1.5; % same here too 
dist = makedist('weibull','a',T,'b',b,'c',t0)
[h,p,adstat,cv] = adtest(data,'Distribution',dist)

[VBBV]

Standard matlab weibull only allows 2 parameter dist as Jeff said But you can use mle to fit a custom distribution Here is the resource below how to do it

http://www.mathworks.com/help/stats/mle.html#btt7ilj