p_of_hit=Uniform(0,1)
p_of_hit
binomial=Binomial(7,p_of_hit)
Copy_of_p_of_hit=Uniform(0,1)
Copy_of_p_of_hit
Copy_of_binomial=Binomial(7,Copy_of_p_of_hit)
0.5 0.5
0.5 0.5
0.5 0.5
0.5 0.5
0.5 0.5
0.5 0.5
0.5 0.5
Copy_of_Node5 Node5 Copy_2_of_Node5 Copy_3_of_Node5 Copy_6_of_Node5 Copy_5_of_Node5 Copy_4_of_Node5 Copy_of_binomial
Node1=Node5+Copy_2_of_Node5+Copy_3_of_Node5+Copy_4_of_Node5+Copy_5_of_Node5+Copy_6_of_Node5+Copy_of_Node5+Copy_of_binomial
Copy_2_of_p_of_hit=Uniform(0,1)
Copy_2_of_p_of_hit
Copy_2_of_binomial=Binomial(7,Copy_2_of_p_of_hit)
Copy_2_of_binomial
1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1
p of hit
118 199 167 229
Observed hits
118 479 172 512
Example A: the node "observed hits" is abinomial function which accepts evidence (0 to 7), the node "p of hit" (unifrom function) returns the most probable probabilities of hit given the number of hits observed.
37 558 224 642
p of hit
494 200 543 230
Observed hits
766 239 820 272
Example B: Everything similar to Example A, but I want to progressively offer evidence to the binomial node. The seven observers can "hit" or "miss", they are added up in the node "number of hits, that is the output of the binomial node... therefore by setting all the observers to miss, the "number of hits" node equals 0 (it works... so far). But its value is not used as evidence for upgrading the binomial "observed hits", and the "p of hit"
927 151 1122 319
Observer 6
1020 454 1068 484
Observer 7
969 377 1017 407
Observer 5
1013 539 1061 569
Observer 4
861 577 909 607
Observer 1
500 449 548 479
Observr 2
494 533 542 563
Observer 3
676 578 724 608
Number of hits
727 423 843 453
Copy of p of hit
399 826 448 856
Copy of Observed hits
767 783 821 816
Observation
1060 795 1131 839