function ft_realtime_downsample(cfg)

% FT_REALTIME_DOWNSAMPLE reads realtime data from one buffer and writes it after downsampling

% to another buffer.

%

% Use as

% ft_realtime_downsample(cfg)

% with the following configuration options

% cfg.channel = cell-array, see FT_CHANNELSELECTION (default = 'all')

% cfg.decimation = integer, downsampling factor (default = 1, no downsampling)

% cfg.order = interger, order of butterworth lowpass filter (default = 4)

% cfg.cutoff = double, cutoff frequency of lowpass filter (default = 0.8*Nyquist-freq.)

%

% The source of the data is configured as

% cfg.source.dataset = string

% or alternatively to obtain more low-level control as

% cfg.source.datafile = string

% cfg.source.headerfile = string

% cfg.source.eventfile = string

% cfg.source.dataformat = string, default is determined automatic

% cfg.source.headerformat = string, default is determined automatic

% cfg.source.eventformat = string, default is determined automatic

%

% The target to write the data to is configured as

% cfg.target.datafile = string, target destination for the data (default = 'buffer://localhost:1972')

% cfg.target.dataformat = string, default is determined automatic

%

% To stop this realtime function, you have to press Ctrl-C

% Copyright (C) 2008, Robert Oostenveld

% Copyright (C) 2010, Stefan Klanke

%

% This file is part of FieldTrip, see http://www.fieldtriptoolbox.org

% for the documentation and details.

%

% FieldTrip is 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.

%

% FieldTrip 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.

%

% You should have received a copy of the GNU General Public License

% along with FieldTrip. If not, see <http://www.gnu.org/licenses/>.

%

% $Id$

% set the defaults

if ~isfield(cfg, 'source'), cfg.source = []; end

if ~isfield(cfg, 'target'), cfg.target = []; end

if ~isfield(cfg.source, 'headerformat'), cfg.source.headerformat = []; end % default is detected automatically

if ~isfield(cfg.source, 'dataformat'), cfg.source.dataformat = []; end % default is detected automatically

if ~isfield(cfg.target, 'headerformat'), cfg.target.headerformat = []; end % default is detected automatically

if ~isfield(cfg.target, 'dataformat'), cfg.target.dataformat = []; end % default is detected automatically

if ~isfield(cfg.target, 'datafile'), cfg.target.datafile = 'buffer://localhost:1972'; end

if ~isfield(cfg, 'minblocksize'), cfg.minblocksize = 0; end % in seconds

if ~isfield(cfg, 'maxblocksize'), cfg.maxblocksize = 1; end % in seconds

if ~isfield(cfg, 'channel'), cfg.channel = 'all'; end

if ~isfield(cfg, 'decimation'), cfg.decimation = 1; end

if ~isfield(cfg, 'order'), cfg.order = 4; end

if cfg.decimation < 1 || cfg.decimation~=round(cfg.decimation)

error 'Decimation factor must be integer quantity >= 1';

end

% translate dataset into datafile+headerfile

cfg.source = ft_checkconfig(cfg.source, 'dataset2files', 'yes');

cfg.target = ft_checkconfig(cfg.target, 'dataset2files', 'yes');

ft_checkconfig(cfg.source, 'required', {'datafile' 'headerfile'});

ft_checkconfig(cfg.target, 'required', {'datafile' 'headerfile'});

% ensure that the persistent variables related to caching are cleared

clear ft_read_header

% read the header for the first time

hdr = ft_read_header(cfg.source.headerfile);

fprintf('updating the header information, %d samples available\n', hdr.nSamples*hdr.nTrials);

targethdr = hdr;

targethdr.Fs = targethdr.Fs/cfg.decimation;

% Calculate cutoff frequency expressed in Nyquist freq of original signal

if ~isfield(cfg, 'cutoff')

cutoff = 0.8/cfg.decimation;

else

cutoff = cfg.cutoff / (0.5*hdr.Fs);

if cutoff >= 1

ft_error('Cutoff frequency too high');

end

end

[B,A] = butter(cfg.order, cutoff, 'low');

% define a subset of channels for reading

cfg.channel = ft_channelselection(cfg.channel, hdr.label);

chanindx = match_str(hdr.label, cfg.channel);

nchan = length(chanindx);

if nchan==0

ft_error('no channels were selected');

end

minblocksmp = round(cfg.minblocksize*hdr.Fs);

minblocksmp = max(minblocksmp, 1);

maxblocksmp = round(cfg.maxblocksize*hdr.Fs);

prevSample = 0;

count = 0;

% create filter model and downsampling model

FM = online_filter_init(B, A, zeros(nchan, 1));

DM = online_downsample_init(cfg.decimation);

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

% this is the general BCI loop where realtime incoming data is handled

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

while true

% determine number of samples available in buffer

hdr = ft_read_header(cfg.source.headerfile, 'cache', true);

% see whether new samples are available

newsamples = (hdr.nSamples*hdr.nTrials-prevSample);

if newsamples>=minblocksmp

% determine the samples to copy from source to target

begsample = prevSample+1;

endsample = prevSample + min(newsamples, maxblocksmp);

% remember up to where the data was read

prevSample = endsample;

count = count + 1;

fprintf('processing segment %d from sample %d to %d\n', count, begsample, endsample);

% read data segment

dat = ft_read_data(cfg.source.datafile, 'dataformat', cfg.source.dataformat, 'begsample', begsample, 'endsample', endsample, 'chanindx', chanindx, 'checkboundary', false);

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

% from here onward it is specific to the downsampling

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

[FM, dat] = online_filter_apply(FM, dat);

[DM, xd] = online_downsample_apply(DM, dat);

if count==1

% flush the file, write the header and subsequently write the data segment

ft_write_data(cfg.target.datafile, dat, 'header', targethdr, 'dataformat', cfg.target.dataformat, 'chanindx', chanindx, 'append', false);

else

% write the data segment

ft_write_data(cfg.target.datafile, dat, 'header', targethdr, 'dataformat', cfg.target.dataformat, 'chanindx', chanindx, 'append', true);

end

end % if enough new samples

end % while true