Mercurial > hg > camir-aes2014

annotate toolboxes/MIRtoolbox1.3.2/somtoolbox/som_read_data.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 sData = som_read_data(filename, varargin)
wolffd@0 2
wolffd@0 3 %SOM_READ_DATA Read data from an ascii file in SOM_PAK format.
wolffd@0 4 %
wolffd@0 5 % sD = som_read_data(filename, dim, [missing])
wolffd@0 6 % sD = som_read_data(filename, [missing])
wolffd@0 7 %
wolffd@0 8 % sD = som_read_data('system.data');
wolffd@0 9 % sD = som_read_data('system.data',10);
wolffd@0 10 % sD = som_read_data('system.data','*');
wolffd@0 11 % sD = som_read_data('system.data',10,'*');
wolffd@0 12 %
wolffd@0 13 % Input and output arguments ([]'s are optional):
wolffd@0 14 % filename (string) input file name
wolffd@0 15 % dim (scalar) input space dimension
wolffd@0 16 % [missing] (string) string which indicates a missing component
wolffd@0 17 % value, 'NaN' by default
wolffd@0 18 %
wolffd@0 19 % sD (struct) data struct
wolffd@0 20 %
wolffd@0 21 % Reads data from an ascii file. The file must be in SOM_PAK format,
wolffd@0 22 % except that it may lack the input space dimension from the first
wolffd@0 23 % line.
wolffd@0 24 %
wolffd@0 25 % For more help, try 'type som_read_data' or check out online documentation.
wolffd@0 26 % See also SOM_WRITE_DATA, SOM_READ_COD, SOM_WRITE_COD, SOM_DATA_STRUCT.
wolffd@0 27
wolffd@0 28 %%%%%%%%%%%%% DETAILED DESCRIPTION %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
wolffd@0 29 %
wolffd@0 30 % som_read_data
wolffd@0 31 %
wolffd@0 32 % PURPOSE
wolffd@0 33 %
wolffd@0 34 % Reads data from an ascii file in SOM_PAK format.
wolffd@0 35 %
wolffd@0 36 % SYNTAX
wolffd@0 37 %
wolffd@0 38 % sD = som_read_data(filename)
wolffd@0 39 % sD = som_read_data(..., dim)
wolffd@0 40 % sD = som_read_data(..., 'missing')
wolffd@0 41 % sD = som_read_data(..., dim, 'missing')
wolffd@0 42 %
wolffd@0 43 % DESCRIPTION
wolffd@0 44 %
wolffd@0 45 % This function is offered for compatibility with SOM_PAK, a SOM software
wolffd@0 46 % package in C. It reads data from a file in SOM_PAK format.
wolffd@0 47 %
wolffd@0 48 % The SOM_PAK data file format is as follows. The first line must
wolffd@0 49 % contain the input space dimension and nothing else. The following
wolffd@0 50 % lines are comment lines, empty lines or data lines. Unlike programs
wolffd@0 51 % in SOM_PAK, this function can also determine the input dimension
wolffd@0 52 % from the first data lines, if the input space dimension line is
wolffd@0 53 % missing. Note that the SOM_PAK format is not fully supported: data
wolffd@0 54 % vector 'weight' and 'fixed' properties are ignored (they are treated
wolffd@0 55 % as labels).
wolffd@0 56 %
wolffd@0 57 % Each data line contains one data vector and its labels. From the beginning
wolffd@0 58 % of the line, first are values of the vector components separated by
wolffd@0 59 % whitespaces, then labels also separated by whitespaces. If there are
wolffd@0 60 % missing values in the vector, the missing value marker needs to be
wolffd@0 61 % specified as the last input argument ('NaN' by default). The missing
wolffd@0 62 % values are stored as NaNs in the data struct.
wolffd@0 63 %
wolffd@0 64 % Comment lines start with '#'. Comment lines as well as empty lines are
wolffd@0 65 % ignored, except if the comment lines that start with '#n' or '#l'. In that
wolffd@0 66 % case the line should contain names of the vector components or label names
wolffd@0 67 % separated by whitespaces.
wolffd@0 68 %
wolffd@0 69 % NOTE: The minimum value Matlab is able to deal with (realmax)
wolffd@0 70 % should not appear in the input file. This is because function sscanf is
wolffd@0 71 % not able to read NaNs: the NaNs are in the read phase converted to value
wolffd@0 72 % realmax.
wolffd@0 73 %
wolffd@0 74 % REQUIRED INPUT ARGUMENTS
wolffd@0 75 %
wolffd@0 76 % filename (string) input filename
wolffd@0 77 %
wolffd@0 78 % OPTIONAL INPUT ARGUMENTS
wolffd@0 79 %
wolffd@0 80 % dim (scalar) input space dimension
wolffd@0 81 % missing (string) string used to denote missing components (NaNs);
wolffd@0 82 % default is 'NaN'
wolffd@0 83 %
wolffd@0 84 % OUTPUT ARGUMENTS
wolffd@0 85 %
wolffd@0 86 % sD (struct) the resulting data struct
wolffd@0 87 %
wolffd@0 88 % EXAMPLES
wolffd@0 89 %
wolffd@0 90 % The basic usage is:
wolffd@0 91 % sD = som_read_data('system.data');
wolffd@0 92 %
wolffd@0 93 % If you know the input space dimension beforehand, and the file does
wolffd@0 94 % not contain it on the first line, it helps if you specify it as the
wolffd@0 95 % second argument:
wolffd@0 96 % sD = som_read_data('system.data',9);
wolffd@0 97 %
wolffd@0 98 % If the missing components in the data are marked with some other
wolffd@0 99 % characters than with 'NaN', you can specify it with the last argument:
wolffd@0 100 % sD = som_read_data('system.data',9,'*')
wolffd@0 101 % sD = som_read_data('system.data','NaN')
wolffd@0 102 %
wolffd@0 103 % Here's an example data file:
wolffd@0 104 %
wolffd@0 105 % 5
wolffd@0 106 % #n one two three four five
wolffd@0 107 % #l ID
wolffd@0 108 % 10 2 3 4 5 1stline label
wolffd@0 109 % 0.4 0.3 0.2 0.5 0.1 2ndline label1 label2
wolffd@0 110 % # comment line: missing components are indicated by 'x':s
wolffd@0 111 % 1 x 1 x 1 3rdline missing_components
wolffd@0 112 % x 1 2 2 2
wolffd@0 113 % x x x x x 5thline emptyline
wolffd@0 114 %
wolffd@0 115 % SEE ALSO
wolffd@0 116 %
wolffd@0 117 % som_write_data Writes data structs/matrices to a file in SOM_PAK format.
wolffd@0 118 % som_read_cod Read a map from a file in SOM_PAK format.
wolffd@0 119 % som_write_cod Writes data struct into a file in SOM_PAK format.
wolffd@0 120 % som_data_struct Creates data structs.
wolffd@0 121
wolffd@0 122 % Copyright (c) 1997-2000 by the SOM toolbox programming team.
wolffd@0 123 % http://www.cis.hut.fi/projects/somtoolbox/
wolffd@0 124
wolffd@0 125 % Version 1.0beta ecco 221097
wolffd@0 126 % Version 2.0beta ecco 060899, juuso 151199
wolffd@0 127
wolffd@0 128 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
wolffd@0 129 %% check arguments
wolffd@0 130
wolffd@0 131 error(nargchk(1, 3, nargin)) % check no. of input args is correct
wolffd@0 132
wolffd@0 133 dont_care = 'NaN'; % default don't care string
wolffd@0 134 comment_start = '#'; % the char a SOM_PAK command line starts with
wolffd@0 135 comp_name_line = '#n'; % string denoting a special command line,
wolffd@0 136 % which contains names of each component
wolffd@0 137 label_name_line = '#l'; % string denoting a special command line,
wolffd@0 138 % which contains names of each label
wolffd@0 139 block_size = 1000; % block size used in file read
wolffd@0 140
wolffd@0 141 kludge = num2str(realmax, 100); % used in sscanf
wolffd@0 142
wolffd@0 143
wolffd@0 144 % open input file
wolffd@0 145
wolffd@0 146 fid = fopen(filename);
wolffd@0 147 if fid < 0
wolffd@0 148 error(['Cannot open ' filename]);
wolffd@0 149 end
wolffd@0 150
wolffd@0 151 % process input arguments
wolffd@0 152
wolffd@0 153 if nargin == 2
wolffd@0 154 if isstr(varargin{1})
wolffd@0 155 dont_care = varargin{1};
wolffd@0 156 else
wolffd@0 157 dim = varargin{1};
wolffd@0 158 end
wolffd@0 159 elseif nargin == 3
wolffd@0 160 dim = varargin{1};
wolffd@0 161 dont_care = varargin{2};
wolffd@0 162 end
wolffd@0 163
wolffd@0 164 % if the data dimension is not specified, find out what it is
wolffd@0 165
wolffd@0 166 if nargin == 1 | (nargin == 2 & isstr(varargin{1}))
wolffd@0 167
wolffd@0 168 fpos1 = ftell(fid); c1 = 0; % read first non-comment line
wolffd@0 169 while c1 == 0,
wolffd@0 170 line1 = strrep(fgetl(fid), dont_care, kludge);
wolffd@0 171 [l1, c1] = sscanf(line1, '%f ');
wolffd@0 172 end
wolffd@0 173
wolffd@0 174 fpos2 = ftell(fid); c2 = 0; % read second non-comment line
wolffd@0 175 while c2 == 0,
wolffd@0 176 line2 = strrep(fgetl(fid), dont_care, kludge);
wolffd@0 177 [l2, c2] = sscanf(line2, '%f ');
wolffd@0 178 end
wolffd@0 179
wolffd@0 180 if (c1 == 1 & c2 ~= 1) | (c1 == c2 & c1 == 1 & l1 == 1)
wolffd@0 181 dim = l1;
wolffd@0 182 fseek(fid, fpos2, -1);
wolffd@0 183 elseif (c1 == c2)
wolffd@0 184 dim = c1;
wolffd@0 185 fseek(fid, fpos1, -1);
wolffd@0 186 warning on
wolffd@0 187 warning(['Automatically determined data dimension is ' ...
wolffd@0 188 num2str(dim) '. Is it correct?']);
wolffd@0 189 else
wolffd@0 190 error(['Invalid header line: ' line1]);
wolffd@0 191 end
wolffd@0 192 end
wolffd@0 193
wolffd@0 194 % check the dimension is valid
wolffd@0 195
wolffd@0 196 if dim < 1 | dim ~= round(dim)
wolffd@0 197 error(['Illegal data dimension: ' num2str(dim)]);
wolffd@0 198 end
wolffd@0 199
wolffd@0 200 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
wolffd@0 201 %% read data
wolffd@0 202
wolffd@0 203 sData = som_data_struct(zeros(1, dim), 'name', filename);
wolffd@0 204 lnum = 0; % data vector counter
wolffd@0 205 data_temp = zeros(block_size, dim);
wolffd@0 206 labs_temp = cell(block_size, 1);
wolffd@0 207 comp_names = sData.comp_names;
wolffd@0 208 label_names = sData.label_names;
wolffd@0 209 form = [repmat('%g',[1 dim-1]) '%g%[^ \t]'];
wolffd@0 210
wolffd@0 211 limit = block_size;
wolffd@0 212 while 1,
wolffd@0 213 li = fgetl(fid); % read next line
wolffd@0 214 if ~isstr(li), break, end; % is this the end of file?
wolffd@0 215
wolffd@0 216 % all missing vectors are replaced by value realmax because
wolffd@0 217 % sscanf is not able to read NaNs
wolffd@0 218 li = strrep(li, dont_care, kludge);
wolffd@0 219 [data, c, err, n] = sscanf(li, form);
wolffd@0 220 if c < dim % if there were less numbers than dim on the input file line
wolffd@0 221 if c == 0
wolffd@0 222 if strncmp(li, comp_name_line, 2) % component name line?
wolffd@0 223 li = strrep(li(3:end), kludge, dont_care); i = 0; c = 1;
wolffd@0 224 while c
wolffd@0 225 [s, c, e, n] = sscanf(li, '%s%[^ \t]');
wolffd@0 226 if ~isempty(s), i = i + 1; comp_names{i} = s; li = li(n:end); end
wolffd@0 227 end
wolffd@0 228
wolffd@0 229 if i ~= dim
wolffd@0 230 error(['Illegal number of component names: ' num2str(i) ...
wolffd@0 231 ' (dimension is ' num2str(dim) ')']);
wolffd@0 232 end
wolffd@0 233 elseif strncmp(li, label_name_line, 2) % label name line?
wolffd@0 234 li = strrep(li(3:end), kludge, dont_care); i = 0; c = 1;
wolffd@0 235 while c
wolffd@0 236 [s, c, e, n] = sscanf(li, '%s%[^ \t]');
wolffd@0 237 if ~isempty(s), i = i + 1; label_names{i} = s; li = li(n:end); end
wolffd@0 238 end
wolffd@0 239 elseif ~strncmp(li, comment_start, 1) % not a comment, is it error?
wolffd@0 240 [s, c, e, n] = sscanf(li, '%s%[^ \t]');
wolffd@0 241 if c
wolffd@0 242 error(['Invalid vector on input file data line ' ...
wolffd@0 243 num2str(lnum+1) ': [' deblank(li) ']']),
wolffd@0 244 end
wolffd@0 245 end
wolffd@0 246 else
wolffd@0 247 error(['Only ' num2str(c) ' vector components on input file data line ' ...
wolffd@0 248 num2str(lnum+1) ' (dimension is ' num2str(dim) ')']);
wolffd@0 249 end
wolffd@0 250
wolffd@0 251 else
wolffd@0 252
wolffd@0 253 lnum = lnum + 1; % this was a line containing data vector
wolffd@0 254 data_temp(lnum, 1:dim) = data'; % add data to struct
wolffd@0 255
wolffd@0 256 if lnum == limit % reserve more memory if necessary
wolffd@0 257 data_temp(lnum+1:lnum+block_size, 1:dim) = zeros(block_size, dim);
wolffd@0 258 [dummy nl] = size(labs_temp);
wolffd@0 259 labs_temp(lnum+1:lnum+block_size,1:nl) = cell(block_size, nl);
wolffd@0 260 limit = limit + block_size;
wolffd@0 261 end
wolffd@0 262
wolffd@0 263 % read labels
wolffd@0 264
wolffd@0 265 if n < length(li)
wolffd@0 266 li = strrep(li(n:end), kludge, dont_care); i = 0; n = 1; c = 1;
wolffd@0 267 while c
wolffd@0 268 [s, c, e, n_new] = sscanf(li(n:end), '%s%[^ \t]');
wolffd@0 269 if c, i = i + 1; labs_temp{lnum, i} = s; n = n + n_new - 1; end
wolffd@0 270 end
wolffd@0 271 end
wolffd@0 272 end
wolffd@0 273 end
wolffd@0 274
wolffd@0 275 % close input file
wolffd@0 276 if fclose(fid) < 0, error(['Cannot close file ' filename]);
wolffd@0 277 else fprintf(2, '\rdata read ok \n'); end
wolffd@0 278
wolffd@0 279 % set values
wolffd@0 280 data_temp(data_temp == realmax) = NaN;
wolffd@0 281 sData.data = data_temp(1:lnum,:);
wolffd@0 282 sData.labels = labs_temp(1:lnum,:);
wolffd@0 283 sData.comp_names = comp_names;
wolffd@0 284 sData.label_names = label_names;
wolffd@0 285
wolffd@0 286 return;
wolffd@0 287
wolffd@0 288 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%