nt_destep

[y,stepList]=nt_destep(x,thresh,guard,depth,minstep) - remove step glitch from MEG data

  y: step-removed data
  stepList: indices of steps

  x: data to clean (time * channels)
  thresh: threshold for variance reduction [default: 0.1]
  guard: minimum duration of stable interval in samples [default: 1000]
  depth: recursion depth for nt_split [default:6], determines number of steps
  minstep: if step size smaller skip step

 Searches for large steps, removes them if variance ratio less than 
 "thresh", and both intervals shorter than "guard".

 See also nt_detrend, nt_deglitch, nt_split.

 NoiseTools.

0001 function [y,stepList]=nt_destep(x,thresh,guard,depth,minstep);
0002 %[y,stepList]=nt_destep(x,thresh,guard,depth,minstep) - remove step glitch from MEG data
0003 %
0004 %  y: step-removed data
0005 %  stepList: indices of steps
0006 %
0007 %  x: data to clean (time * channels)
0008 %  thresh: threshold for variance reduction [default: 0.1]
0009 %  guard: minimum duration of stable interval in samples [default: 1000]
0010 %  depth: recursion depth for nt_split [default:6], determines number of steps
0011 %  minstep: if step size smaller skip step
0012 %
0013 % Searches for large steps, removes them if variance ratio less than
0014 % "thresh", and both intervals shorter than "guard".
0015 %
0016 % See also nt_detrend, nt_deglitch, nt_split.
0017 %
0018 % NoiseTools.
0019 nt_greetings;
0020 
0021 if nargin<1; error('!'); end
0022 if nargin<2||isempty(thresh); thresh=0.1; end
0023 if nargin<3||isempty(guard); guard=100; end
0024 if nargin<4 || isempty(depth); depth=6; end
0025 if nargin<5 ; minstep=[]; end
0026 
0027 if isempty(minstep); minstep=(max(x(:))-min(x(:)))*0.0000001; end
0028 y=x;
0029 
0030 %disp(numel(thresh))
0031 
0032 for iChan=1:size(x,2)
0033     % find step indices
0034     [stepList,score_vector,score]=nt_split(x(:,iChan),depth,thresh,guard,minstep);
0035     
0036     if ~isempty(stepList)
0037         stepList=[0,stepList,size(x,1)];
0038         for iSplit=2:numel(stepList)-1
0039             y1=y(stepList(iSplit-1)+1:stepList(iSplit),iChan); % plateau before
0040             y2=y(stepList(iSplit)+1:stepList(iSplit+1),iChan); % plateau after
0041             step=(mean(y2)-mean(y1));
0042             y(stepList(iSplit)+1:end,iChan)=y(stepList(iSplit)+1:end,iChan)-step;
0043         end
0044     end
0045 end
0046 stepList=stepList(2:end-1);
0047 
0048 
0049 if ~nargout
0050     % don't return values, just plot
0051     disp(['steps at: ', num2str(stepList(2:end-1))]);
0052     figure(1); clf; nt_banner('nt_destep');
0053     xx=[x(:),y(:)];
0054     lim=[min(xx(:)),max(xx(:))]; lim=[lim(1)-0.1*(lim(2)-lim(1)), lim(2)+0.1*(lim(2)-lim(1))];
0055     subplot 211; plot([x,y]); xlabel('samples'); ylim(lim); legend('raw','clean'); legend boxoff
0056     subplot 212;     plot(y,'r'); xlabel('samples'); legend('clean'); legend boxoff
0057     clear y, stepList;
0058 end
0059 
0060 
0061 
0062 % test code
0063 % if 0
0064 %     N=8;
0065 %     Wn=0.25; % CTF corner frequency is 1/4 of sampling rate
0066 %     nsamples=100;
0067 %     [B,A] = butter(N,Wn);
0068 %     y=filter(B,A,ones(nsamples,1));
0069 %     BETA0=[1, zeros(1,8), 1, zeros(1,8)];
0070 %     fun = @(beta,x)(filter([beta(1),beta(2),beta(3),beta(4),beta(5),beta(6),beta(7),beta(8),beta(9)],...
0071 %         [beta(10),beta(11),beta(12),beta(13),beta(14),beta(15),beta(16),beta(17),beta(18)],x));
0072 %     x=ones(nsamples,1);
0073 %     BETA = nlinfit(x,y,fun,BETA0);
0074 %     BETA=BETA/BETA(10);
0075 %     BB=BETA(1:9); AA=BETA(10:end);
0076 %     figure(1); clf;
0077 %     plot([y,filter(BB,AA,ones(size(y)))])
0078 % end
0079 if 0
0080     N=8;
0081     Wn=0.25; % CTF corner frequency is 1/4 of sampling rate
0082     [B,A] = butter(N,Wn);
0083     x=[zeros(1100,1);ones(2000,1)];
0084     x=filter(B,A,x);
0085     x=x+0.01*randn(size(x));
0086     [y,stepList]=nt_destep(x);
0087     figure(1); clf;
0088     subplot 211;
0089     plot([x,y]); legend('raw','destep');legend boxoff
0090     subplot 212; 
0091     plot([y,nt_deglitch(y,stepList)]); 
0092     legend('destep','deglitch');legend boxoff
0093 end
0094 if 0
0095     N=8;
0096     Wn=0.25; % CTF corner frequency is 1/4 of sampling rate
0097     [B,A] = butter(N,Wn);
0098     x=[zeros(1000,1);ones(1000,1);2*ones(1000,1);3*ones(1000,1);4*ones(1000,1);0*ones(1000,1)];
0099     x=filter(B,A,x);
0100     x=x+0.0*randn(size(x));
0101     [y,stepList]=nt_destep(x);
0102     figure(1); clf;
0103     subplot 211;
0104     plot([x,y]); legend('raw','destep'); legend boxoff
0105     subplot 212; 
0106     plot([y,nt_deglitch(y,stepList)]); 
0107     legend('destep','deglitch');legend boxoff
0108 end
0109 if 0
0110     x=ft_read_data('/data/meg/litvak/phantom090715_BrainampDBS_20150709_18.ds'); x=permute(x,[2,1,3]); x=x(:,33:306,:);
0111     x=nt_demean(x(1:2300,100,1));
0112     x=x*10^12; % -->pT
0113     [y,stepList]=nt_destep(x);
0114     figure(1); clf;
0115     subplot 211;
0116     plot([x,y]); legend('raw','destep'); legend boxoff
0117     subplot 212; 
0118     plot([y,nt_deglitch(y,stepList)]); 
0119     legend('destep','deglitch');legend boxoff
0120 end
0121 
0122