Mercurial > hg > camir-aes2014

annotate toolboxes/MIRtoolbox1.3.2/somtoolbox/som_read_cod.m @ 0:e9a9cd732c1e tip

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
first hg version after svn
author wolffd
date Tue, 10 Feb 2015 15:05:51 +0000
parents
children
rev   line source
wolffd@0 1 function sMap = som_read_cod(filename)
wolffd@0 2
wolffd@0 3 %SOM_READ_COD Reads a SOM_PAK format codebook file.
wolffd@0 4 %
wolffd@0 5 % sMap = som_read_cod(filename);
wolffd@0 6 %
wolffd@0 7 % sMap = som_read_cod('map1.cod');
wolffd@0 8 %
wolffd@0 9 % Input and output arguments:
wolffd@0 10 % filename (string) name of input file
wolffd@0 11 % sMap (struct) self-organizing map structure
wolffd@0 12 %
wolffd@0 13 % The file must be in SOM_PAK format. Empty lines and lines starting
wolffd@0 14 % with a '#' are ignored, except the ones starting with '#n'. The strings
wolffd@0 15 % after '#n' are read to field 'comp_names' of the map structure.
wolffd@0 16 %
wolffd@0 17 % For more help, try 'type som_read_cod' or check out online documentation.
wolffd@0 18 % See also SOM_WRITE_COD, SOM_READ_DATA, SOM_WRITE_DATA, SOM_MAP_STRUCT.
wolffd@0 19
wolffd@0 20 %%%%%%%%%%%%% DETAILED DESCRIPTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
wolffd@0 21 %
wolffd@0 22 % som_read_cod
wolffd@0 23 %
wolffd@0 24 % PURPOSE
wolffd@0 25 %
wolffd@0 26 % Reads a Self-Organizing Map from an ascii file in SOM_PAK format.
wolffd@0 27 %
wolffd@0 28 % SYNTAX
wolffd@0 29 %
wolffd@0 30 % sMap = som_read_cod(filename);
wolffd@0 31 %
wolffd@0 32 % DESCRIPTION
wolffd@0 33 %
wolffd@0 34 % This function is offered for compatibility with SOM_PAK, a SOM
wolffd@0 35 % software package in C. It reads map files written in SOM_PAK format.
wolffd@0 36 %
wolffd@0 37 % The SOM_PAK map file format is as follows. The first line must contain
wolffd@0 38 % the input space dimension, lattice type ('rect' or 'hexa'), map grid
wolffd@0 39 % size in x-direction, map grid size in y-direction, and neighborhood
wolffd@0 40 % function ('bubble' or 'gaussian'), in that order. The following lines
wolffd@0 41 % are comment lines, empty lines or data lines.
wolffd@0 42 %
wolffd@0 43 % Each data line contains the weight vector of one map unit and its
wolffd@0 44 % labels. From the beginning of the line, first are values of the vector
wolffd@0 45 % components separated by whitespaces, then labels, again separated by
wolffd@0 46 % whitespaces. The order of map units in the file are one row at a time
wolffd@0 47 % from right to left, from the top to the bottom of the map (x-direction
wolffd@0 48 % first, then y-direction).
wolffd@0 49 %
wolffd@0 50 % Comment lines start with '#'. Comment lines as well as empty lines are
wolffd@0 51 % ignored, except if the comment line starts with '#n'. In that case the
wolffd@0 52 % line should contain names of the vector components separated by
wolffd@0 53 % whitespaces.
wolffd@0 54 %
wolffd@0 55 % In the returned map struct, several fields has to be set to default
wolffd@0 56 % values, since the SOM_PAK file does not contain information on
wolffd@0 57 % them. These include map shape ('sheet'), mask ([1 ... 1]),
wolffd@0 58 % normalizations (none), trainhist (two entries, first with algorithm
wolffd@0 59 % 'init' and the second with 'seq', both with data name 'unknown'),
wolffd@0 60 % possibly also component names ('Var1',...).
wolffd@0 61 %
wolffd@0 62 % REQUIRED INPUT PARAMETERS
wolffd@0 63 %
wolffd@0 64 % filename (string) the name of the input file
wolffd@0 65 %
wolffd@0 66 % OUTPUT ARGUMENTS
wolffd@0 67 %
wolffd@0 68 % sMap (struct) the resulting SOM struct
wolffd@0 69 %
wolffd@0 70 % EXAMPLES
wolffd@0 71 %
wolffd@0 72 % sMap = som_read_cod('map1.cod');
wolffd@0 73 %
wolffd@0 74 % SEE ALSO
wolffd@0 75 %
wolffd@0 76 % som_write_cod Writes a map struct into a file in SOM_PAK format.
wolffd@0 77 % som_read_data Reads data from an ascii file.
wolffd@0 78 % som_write_data Writes data struct into a file in SOM_PAK format.
wolffd@0 79 % som_map_struct Creates map structs.
wolffd@0 80
wolffd@0 81 % Copyright (c) 1997-2000 by the SOM toolbox programming team.
wolffd@0 82 % http://www.cis.hut.fi/projects/somtoolbox/
wolffd@0 83
wolffd@0 84 % Version 1.0beta ecco 221097
wolffd@0 85 % Version 2.0beta juuso 151199 250400
wolffd@0 86
wolffd@0 87 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
wolffd@0 88 %% check arguments
wolffd@0 89
wolffd@0 90 error(nargchk(1, 1, nargin)) % check no. of input args is correct
wolffd@0 91
wolffd@0 92 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
wolffd@0 93 %% initialize variables
wolffd@0 94
wolffd@0 95 lnum = 0; % codebook vector counter
wolffd@0 96 comment_start = '#'; % the char a SOM_PAK command line starts with
wolffd@0 97 comp_name_line = '#n'; % string used to start a special command line,
wolffd@0 98 % which contains names of each component
wolffd@0 99
wolffd@0 100 % open input file
wolffd@0 101
wolffd@0 102 fid = fopen(filename);
wolffd@0 103 if fid < 0, error(['Cannot open ' filename]); end
wolffd@0 104
wolffd@0 105 % read header line
wolffd@0 106
wolffd@0 107 ok_cnt = 0;
wolffd@0 108 lin = fgetl(fid); li = lin;
wolffd@0 109 [dim c err n] = sscanf(li, '%d%[^ \t]'); ok_cnt=ok_cnt+c; li = li(n:end);
wolffd@0 110 [lattice c err n] = sscanf(li,'%s%[^ \t]'); ok_cnt=ok_cnt+c; li = li(n:end);
wolffd@0 111 [msize(2) c err n] = sscanf(li, '%d%[^ \t]'); ok_cnt=ok_cnt+c; li = li(n:end);
wolffd@0 112 [msize(1) c err n] = sscanf(li, '%d%[^ \t]'); ok_cnt=ok_cnt+c; li = li(n:end);
wolffd@0 113 [neigh c err n] = sscanf(li, '%s%[^ \t\n]'); ok_cnt=ok_cnt+c;
wolffd@0 114
wolffd@0 115 if ok_cnt ~= 5
wolffd@0 116 error([ 'Invalid header line: ' lin ]);
wolffd@0 117 end
wolffd@0 118
wolffd@0 119 % create map struct and set its fields according to header line
wolffd@0 120
wolffd@0 121 munits = prod(msize);
wolffd@0 122 sMap = som_map_struct(dim, 'msize', msize, ...
wolffd@0 123 lattice, 'sheet', 'neigh', neigh);
wolffd@0 124 [sT0, ok] = som_set('som_train','algorithm','init','data_name','unknown');
wolffd@0 125 sT1 = som_set('som_train','algorithm','seq','data_name','unknown',...
wolffd@0 126 'neigh',neigh,'mask',ones(dim,1));
wolffd@0 127 [sMap, ok, msgs] = som_set(sMap,'name',filename,'trainhist',{sT0,sT1});
wolffd@0 128
wolffd@0 129 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
wolffd@0 130 %% read codebook from the file
wolffd@0 131
wolffd@0 132 codebook = zeros(munits,dim);
wolffd@0 133 labels = cell(munits,1);
wolffd@0 134 comp_names = sMap.comp_names;
wolffd@0 135 form = [repmat('%f',[1 dim-1]) '%f%[^ \t]'];
wolffd@0 136
wolffd@0 137 while 1,
wolffd@0 138 li = fgetl(fid); % read next line
wolffd@0 139 if ~isstr(li), break, end; % is this the end of file?
wolffd@0 140
wolffd@0 141 [data, c, err, n] = sscanf(li, form);
wolffd@0 142 if c < dim % if there were less numbers than dim on the input file line
wolffd@0 143 if c == 0
wolffd@0 144 if strncmp(li, comp_name_line, 2) % component name line?
wolffd@0 145 li = li(3:end); i = 0; c = 1;
wolffd@0 146 while c
wolffd@0 147 [s, c, e, n] = sscanf(li, '%s%[^ \t]');
wolffd@0 148 if ~isempty(s), i = i + 1; comp_names{i} = s; li = li(n:end); end
wolffd@0 149 end
wolffd@0 150
wolffd@0 151 if i ~= dim
wolffd@0 152 error(['Illegal number of component names: ' num2str(i) ...
wolffd@0 153 ' (dimension is ' num2str(dim) ')']);
wolffd@0 154 end
wolffd@0 155 elseif ~strncmp(li, comment_start, 1) % not a comment, is it error?
wolffd@0 156 [s, c, e, n] = sscanf(li, '%s%[^ \t]');
wolffd@0 157 if c
wolffd@0 158 error(['Invalid vector on input file line ' ...
wolffd@0 159 num2str(lnum+1) ': [' deblank(li) ']']),
wolffd@0 160 end
wolffd@0 161 end
wolffd@0 162 else
wolffd@0 163 error(['Only ' num2str(c) ' vector components on input file line ' ...
wolffd@0 164 num2str(lnum+1) ' (dimension is ' num2str(dim) ')']);
wolffd@0 165 end
wolffd@0 166
wolffd@0 167 else
wolffd@0 168
wolffd@0 169 lnum = lnum + 1; % this was a line containing data vector
wolffd@0 170 codebook(lnum, 1:dim) = data'; % add data to struct
wolffd@0 171
wolffd@0 172 % read labels
wolffd@0 173
wolffd@0 174 if n < length(li)
wolffd@0 175 li = li(n:end);
wolffd@0 176 i = 0; n = 1; c = 1;
wolffd@0 177 while c
wolffd@0 178 [s, c, e, n_new] = sscanf(li(n:end), '%s%[^ \t]');
wolffd@0 179 if c, i = i + 1; labels{lnum, i} = s; n = n + n_new - 1; end
wolffd@0 180 end
wolffd@0 181 end
wolffd@0 182 end
wolffd@0 183 end
wolffd@0 184
wolffd@0 185 % close the input file
wolffd@0 186
wolffd@0 187 if fclose(fid) < 0
wolffd@0 188 error(['Cannot close file ' filename]);
wolffd@0 189 else
wolffd@0 190 fprintf(2, '\rmap read ok \n');
wolffd@0 191 end
wolffd@0 192
wolffd@0 193 % check that the number of lines read was correct
wolffd@0 194
wolffd@0 195 if lnum ~= munits
wolffd@0 196 error(['Illegal number of map units: ' num2str(lnum) ' (should be ' num2str(munits) ').']);
wolffd@0 197 end
wolffd@0 198
wolffd@0 199 % set values
wolffd@0 200
wolffd@0 201 % in SOM_PAK the xy-indexing is used, while in Matlab ij-indexing
wolffd@0 202 % therefore, the codebook vectors have to be reorganized
wolffd@0 203
wolffd@0 204 order = reshape([1:munits],msize);
wolffd@0 205 order = reshape(order',[munits 1]);
wolffd@0 206 codebook(order,:) = codebook;
wolffd@0 207 labels(order,:) = labels;
wolffd@0 208
wolffd@0 209 sMap.codebook = codebook;
wolffd@0 210 sMap.labels = labels;
wolffd@0 211 sMap.comp_names = comp_names;
wolffd@0 212
wolffd@0 213 return;
wolffd@0 214
wolffd@0 215 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%