|
wolffd@0
|
1 function [bnt_data, old_values] = transform_data_into_bnt_format(data,cnodes)
|
|
wolffd@0
|
2 % TRANSFORM_DATA_TO_BNT_FORMAT Ensures discrete variables have values 1,2,..,k
|
|
wolffd@0
|
3 % e.g., if the values of a discrete are [0 1 6], they must be mapped to [1 2 3]
|
|
wolffd@0
|
4 %
|
|
wolffd@0
|
5 % data(i,j) is the value for i-th node in j-th case.
|
|
wolffd@0
|
6 % bnt_data(i,j) is the new value.
|
|
wolffd@0
|
7 % old_values{i} are the original values for node i.
|
|
wolffd@0
|
8 % cnodes is the list of all continous nodes, e.g. [3 5] means the 3rd and 5th node is continuous
|
|
wolffd@0
|
9 %
|
|
wolffd@0
|
10 % Author: yimin.zhang@intel.com
|
|
wolffd@0
|
11 % Last updated: Jan. 22, 2002 by Kevin Murphy.
|
|
wolffd@0
|
12
|
|
wolffd@0
|
13 num_nodes=size(data,1);
|
|
wolffd@0
|
14 num_cases=size(data,2);
|
|
wolffd@0
|
15 old_values=cell(1,num_nodes);
|
|
wolffd@0
|
16
|
|
wolffd@0
|
17 for i=1:num_nodes
|
|
wolffd@0
|
18 if (myismember(i,cnodes)==1) %cts nodes no need to be transformed
|
|
wolffd@0
|
19 %just copy the data
|
|
wolffd@0
|
20 bnt_data(i,:)=data(i,:);
|
|
wolffd@0
|
21 continue;
|
|
wolffd@0
|
22 end
|
|
wolffd@0
|
23 values = data(i,:);
|
|
wolffd@0
|
24 sort_v = sort(values);
|
|
wolffd@0
|
25 %remove the duplicate values in sort_v
|
|
wolffd@0
|
26 v_set = unique(sort_v);
|
|
wolffd@0
|
27
|
|
wolffd@0
|
28 %transform the values
|
|
wolffd@0
|
29 for j=1:size(values,2)
|
|
wolffd@0
|
30 index = binary_search(v_set,values(j));
|
|
wolffd@0
|
31 if (index==-1)
|
|
wolffd@0
|
32 fprintf('value not found in tranforming data to bnt format.\n');
|
|
wolffd@0
|
33 return;
|
|
wolffd@0
|
34 end
|
|
wolffd@0
|
35 bnt_data(i,j)=index;
|
|
wolffd@0
|
36 end
|
|
wolffd@0
|
37 old_values{i}=v_set;
|
|
wolffd@0
|
38 end
|
|
wolffd@0
|
39
|
|
wolffd@0
|
40
|
|
wolffd@0
|
41 %%%%%%%%%%%%
|
|
wolffd@0
|
42
|
|
wolffd@0
|
43 function index=binary_search(vector, value)
|
|
wolffd@0
|
44 % BI_SEARCH do binary search for value in the vector
|
|
wolffd@0
|
45 % Author: yimin.zhang@intel.com
|
|
wolffd@0
|
46 % Last updated: Jan. 19, 2002
|
|
wolffd@0
|
47
|
|
wolffd@0
|
48 begin_index=1;
|
|
wolffd@0
|
49 end_index=size(vector,2);
|
|
wolffd@0
|
50 index=-1;
|
|
wolffd@0
|
51 while (begin_index<=end_index)
|
|
wolffd@0
|
52 mid=floor((begin_index+end_index)/2);
|
|
wolffd@0
|
53 if (isstr(vector(mid)))
|
|
wolffd@0
|
54 % need to write a strcmp to return three result (< = >)
|
|
wolffd@0
|
55 else
|
|
wolffd@0
|
56 if (value==vector(mid))
|
|
wolffd@0
|
57 index=mid;
|
|
wolffd@0
|
58 return;
|
|
wolffd@0
|
59 elseif (value>vector(mid))
|
|
wolffd@0
|
60 begin_index=mid+1;
|
|
wolffd@0
|
61 else
|
|
wolffd@0
|
62 end_index=mid-1;
|
|
wolffd@0
|
63 end
|
|
wolffd@0
|
64 end
|
|
wolffd@0
|
65 end
|
|
wolffd@0
|
66 return;
|
