1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
|
function imfcnt = fn_empdfsort(imfcnt,imf_h,imfloor,hbin,ninv)
% imfcnt = fn_empdfsort(imfcnt,imf_h,imfloor,hbin,ninv)
% Variablewise (but multivariate) empirical probability distribution with counts
% sorted into bins variablewise
% Note that since a particular draw (imf_h) can be below "imfloor" and above "imceiling"
% (=(imceiling-imfloor)*hbin), this function allows ninv+2 bins for each variable
%
% imfcnt: initial ninv+2-by-k matrix that records counts in each bin given a column in imfcnt
% if k==1, then only one variable is considered.
% imf_h: k-element vector of particular draws. Need not be a 1-by-k row vector.
% imfloor: k-element vector of low values of imf but imf_h can be below "imfloor" and above "imceiling"
% hbin: k-element vector of bin lengths = (imceilling-imfloor)/ninv.
% ninv: number of bins between "imfloor" and "imceiling"
%------------------
% imfcnt: updated ninv+2-by-k matrix of counts (probability)
%
% January 1999 TZ; Revised, August 2000.
%
% Copyright (C) 1997-2012 Tao Zha
%
% This free software: you can redistribute it and/or modify
% it under the terms of the GNU General Public License as published by
% the Free Software Foundation, either version 3 of the License, or
% (at your option) any later version.
%
% It is distributed in the hope that it will be useful,
% but WITHOUT ANY WARRANTY; without even the implied warranty of
% MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
% GNU General Public License for more details.
%
% If you did not received a copy of the GNU General Public License
% with this software, see <http://www.gnu.org/licenses/>.
%
%** find the bin locations and arrange them to the order of 1, 2, ...
imf_h=imf_h(:); imfloor=imfloor(:); hbin=hbin(:);
countInt = floor( (imf_h-imfloor) ./ hbin ); % k-by-1
% bin locations from <0, 0, 1,..., ninv-1, >=ninv, a total of ninv+2 bins
countInt = 2+countInt'; % 1-by-k row vector, see my shock (1), pp.1
% move everyting by 2 so as to take account of <0
countInt(find(countInt<2)) = 1; % set <0 or -infinity at 1 to start
countInt(find(countInt>=ninv+2)) = ninv+2; % set >=ninv+2 or +infinity at ninv+2 to end
countInt = countInt + (0:length(countInt)-1)*(ninv+2);
% 1-by-k*(ninv+2) index vector to fill the matrix with prob. (counts)
% The term after "+" implies that with every count, we skip ninv+2 to keep
% each column in "imfcnt" with only one element (which is probability)
imfcnt(countInt) = imfcnt(countInt) + 1;
|
