Producing an error vector with a loop

1 view (last 30 days)
I need a loop that varies the 'MinLeaf' value from 200-2000 in steps of 10. All errors should go into one vector called 'LogLike_Error_Vector". Thanks Tamir
Num = xlsread('Data_Class_2015_Student_13.xlsx');
Y = Num(:,end); Y(1)=[];
X = Num(:,1:end-1); X(1,:)=[];
ly = length(Y);
i0 = Y==0;
i1 = Y==1;
im = Y<0;
rand('seed',123)
rnd = rand(ly,1);
iData=rnd>=0 & ~im;
XX = X;
t4=fitrtree(XX(iData,:),Y(iData),'MinLeaf',400);
view(t4,'mode','graph')
t2=prune(t4,'level', 15);
DT = predict(t2,XX);
LogLike_DT_Data = LogLikelihood(DT(iData),Y(iData))/sum(iData)

Accepted Answer

Shoaibur Rahman
Shoaibur Rahman on 1 Jan 2015
Where is your LogLike_Error_Vector? Is it LogLike_DT_Data? If so, you can try using a for loop:
indx = 0;
for MinLeaf = 200:10:2000
indx = indx + 1;
t4=fitrtree(XX(iData,:),Y(iData),'MinLeaf', MinLeaf );
view(t4,'mode','graph') % you can skip this, if not needed
t2=prune(t4,'level', 15);
DT = predict(t2,XX);
LogLike_DT_Data(indx) = LogLikelihood(DT(iData),Y(iData))/sum(iData)
end

More Answers (0)

Categories

Find more on Extend Modeling Environment in Help Center and File Exchange

Tags

Products

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!