Getting 'Reference to non-existent field' Error

I run the code as a m file,but it occurs to me a problem telling Reference to non-existent field...What should I do?Please help me!Thank you very much.

      Reference to non-existent field 'Count(track_id)'.

      Error in getfield (line 37)
          f = s.(deblank(strField)); % deblank field name

      Error in @(x)getfield(x,'Count(track_id)')

      Error in MS (line 70)
      unused, order] =
      sort(arrayfun(@(x)getfield(x,'Count(track_id)'),res(:)),'descend');

Here is my M file code(or you can see it in the attach file):

% set up Million Song paths
global MillionSong MSDsubset
MillionSong ='MillionSongSubset';  % or 'MillionSong' for full set
msd_data_path=[MillionSong,'/data'];
msd_addf_path=[MillionSong,'/AdditionalFiles'];
MSDsubset='subset_'; % or '' for full set
msd_addf_prefix=[msd_addf_path,'/',MSDsubset];
% Check that we can actually read the dataset
assert(exist(msd_data_path,'dir')==7,['msd_data_path ',msd_data_path,' is not found.']);

% path to the Million Song Dataset code
msd_code_path='MSongsDB';
assert(exist(msd_code_path,'dir')==7,['msd_code_path ',msd_code_path,' is wrong.']);
% Build a list of all the files in the dataset
all_files = findAllFiles(msd_data_path);
cnt = length(all_files);
disp(['Number of h5 files found: ',num2str(cnt)]);

% Get info from the first file using our wrapper
h5 = HDF5_Song_File_Reader(all_files{1});
disp(['artist name is: ',h5.get_artist_name()]);
disp([' song title is: ',h5.get_title()]);
methods('HDF5_Song_File_Reader')
%a=h5.get_segments_name;

% Get all artist names by mapping a function to return artist names
% over the cell array of data file names
tic;
all_artist_names = cellfun(@(f) get_artist_name(HDF5_Song_File_Reader(f)), ...
                           all_files, 'UniformOutput', false);
tend = toc;
disp(['All names acquired in ',num2str(tend),' seconds.']);
disp(['First artist name is: ',all_artist_names{1}]);
disp(['There are ',num2str(length(unique(all_artist_names))), ...
      ' unique artist names']);
% takes around 5 min on MacBook Pro to scan 10k files (30ms/file)

% Track metadata database
sqldb = [msd_addf_prefix,'track_metadata.db'];
% Open connection
mksqlite('open',sqldb);

% Run the SQL query.  DISTINCT means we only get the unique names
artist_names = mksqlite('SELECT DISTINCT artist_name FROM songs');
% Close the connection (clean up)
mksqlite('close');
disp(['Found ',num2str(length(artist_names)),' distinct artist names']);
disp('First artist names are:');
for k=1:5;
    disp(artist_names(k).artist_name);
end

mksqlite('open',sqldb);
res = mksqlite(['SELECT DISTINCT artist_id,artist_name,Count(track_id) ' ...
                'FROM songs GROUP BY artist_id']);
mksqlite('close');
disp('Got entries that looks like:');
res(1)

% Sort the results
[unused, order] = sort(arrayfun(@(x)getfield(x,'Count(track_id)'),res(:)),'descend');
res = res(order);
disp('Artist with the most songs is:');
res(1)

Answer 1

[Walter Roberson]

fieldnames have the same rules as hold for variable names: 1 to 63 characters, begin with a Latin alphabetic letter, following characters must be selected from the Latin alphabetic letters, the Latin "arabic digits", or the underscore.

The fieldname 'Count(track_id)' does not follow those rules, as it attempts to use the prohibited characters '(' and ')' as part of the field names.

[Lin Yanzhou]

But the code

mksqlite('open',sqldb); res = mksqlite(['SELECT DISTINCT artist_id,artist_name,Count(track_id) ' ... 'FROM songs GROUP BY artist_id']); mksqlite('close');

also use "track_id" and there is no error about it.