camir-aes2014: toolboxes/FullBNT-1.0.7/bnt/examples/dynamic/mk_mildew

first hg version after svn

rev	line source
wolffd@0	1 function bnet = mk_mildew_dbn()
wolffd@0	2
wolffd@0	3 % DBN for foreacasting the gross yield of wheat based on climatic data,
wolffd@0	4 % observations of leaf area index (LAI) and extension of mildew,
wolffd@0	5 % and knowledge of amount of fungicides used and time of usage.
wolffd@0	6 % From Kjaerulff '95.
wolffd@0	7
wolffd@0	8 Fungi=1; Mildew=2; LAI=3; Precip=4; Temp=5; Micro=6; Solar=7; Photo=8; Dry=9;
wolffd@0	9 n = 9;
wolffd@0	10 intra = zeros(n,n);
wolffd@0	11 intra(Mildew, LAI)=1;
wolffd@0	12 intra(LAI,[Micro Photo])=1;
wolffd@0	13 intra(Precip,Micro)=1;
wolffd@0	14 intra(Temp,[Micro Photo])=1;
wolffd@0	15 intra(Solar,Photo)=1;
wolffd@0	16 intra(Photo,Dry)=1;
wolffd@0	17
wolffd@0	18 inter = zeros(n,n);
wolffd@0	19 inter(Fungi,Mildew)=1;
wolffd@0	20 inter(Mildew,Mildew)=1;
wolffd@0	21 inter(LAI,LAI)=1;
wolffd@0	22 inter(Micro,Mildew)=1;
wolffd@0	23 inter(Dry,Dry)=1;
wolffd@0	24
wolffd@0	25 ns = 2*ones(1,n);
wolffd@0	26 bnet = mk_dbn(intra, inter, ns, 'observed', [Photo]);
wolffd@0	27
wolffd@0	28 for e=1:max(bnet.equiv_class(:))
wolffd@0	29 i = bnet.rep_of_eclass(e);
wolffd@0	30 bnet.CPD{e} = tabular_CPD(bnet,i);
wolffd@0	31 end

Mercurial > hg > camir-aes2014