#climatedata » Tag: Matlab

#climatedata » Tag: Matlab - Recent Posts http://climatedata.blprnt.com/tags/matlab Data Visualization & Climate Reality en-US Sat, 04 Feb 2012 03:21:14 +0000 http://bbpress.org/?v=1.0.2 <![CDATA[Search]]> q http://climatedata.blprnt.com/search.php Arenamontanus on "Matlab" http://climatedata.blprnt.com/topic/matlab#post-53 Wed, 16 Dec 2009 13:08:02 +0000 Arenamontanus 53@http://climatedata.blprnt.com/ Great idea to run them as animations, I have to try that out. I just finished the overall analysis to my own satisfaction, placing it on <a href="http://www.aleph.se/andart/archives/2009/12/significant_anomalies.html" rel="nofollow">http://www.aleph.se/andart/archives/2009/12/significant_anomalies.html</a> Pictures at <a href="http://www.flickr.com/photos/arenamontanus/4186189678/" rel="nofollow">http://www.flickr.com/photos/arenamontanus/4186189678/</a> <a href="http://www.flickr.com/photos/arenamontanus/4189456021/" rel="nofollow">http://www.flickr.com/photos/arenamontanus/4189456021/</a> <a href="http://www.flickr.com/photos/arenamontanus/4189455889/" rel="nofollow">http://www.flickr.com/photos/arenamontanus/4189455889/</a> <a href="http://www.flickr.com/photos/arenamontanus/4189456149/" rel="nofollow">http://www.flickr.com/photos/arenamontanus/4189456149/</a> Overall, Matlab is great for doing calculations and statistics but lousy for actual visualisations. It shouldn't be cumbersome to make a simple colorbar scale showing the right scale, but it is. blprnt on "Matlab" http://climatedata.blprnt.com/topic/matlab#post-50 Wed, 16 Dec 2009 04:51:33 +0000 blprnt 50@http://climatedata.blprnt.com/ Very nice. This series is very effective: <a href="http://www.flickr.com/photos/arenamontanus/4182402460/" title="Suntar by Arenamontanus, on Flickr"><img src="http://farm3.static.flickr.com/2780/4182402460_1acf7a245c.jpg" alt="Suntar" /></a> I'd like to see these as animations - I'd also like to see the boundary lines smoothed out a little bit... Arenamontanus on "Matlab" http://climatedata.blprnt.com/topic/matlab#post-45 Mon, 14 Dec 2009 00:36:11 +0000 Arenamontanus 45@http://climatedata.blprnt.com/ I have placed some of the resulting visualisations in <a href="http://www.flickr.com/photos/arenamontanus/sets/72157622868634041/" rel="nofollow">http://www.flickr.com/photos/arenamontanus/sets/72157622868634041/</a> (and of course the Flickr group). Arenamontanus on "Matlab" http://climatedata.blprnt.com/topic/matlab#post-44 Sun, 13 Dec 2009 22:12:18 +0000 Arenamontanus 44@http://climatedata.blprnt.com/ I wrote a little script to read the data into Matlab for further processing. Nothing particularly clever, mainly a bit of file management and string processing to extract data and convert it into arrays. Some basic visualisations can be done. Right now it just deals with each station separately, but the next step involves merging the data in various ways. Not sure if it works in Octave. <code> % Matlab code to load Land surface climate station records % Assumes the data is unzipped into a subdirectory named Data. % Iterates through the subfolders, loading the data station by station % % Anders Sandberg subdirs=dir('Data'); files={}; llong=[]; llat=[]; hheight=[]; for i=1:length(subdirs) if (subdirs(i).name(1)~='.') datafiles=dir(strcat('Data/',char(subdirs(i).name))); for j=1:length(datafiles) if (datafiles(j).name(1)~='.') filename = strcat('Data/',char(subdirs(i).name),'/',char(datafiles(j).name)); fid=fopen(filename,'r'); %tline = fgetl(fid); %sscanf(tline,'%d'), while (1) tline = fgetl(fid); if (strcmp(tline,'Obs:')) break; end C=textscan(tline,'%s%s', 1,'Delimiter','='); if (strcmp(C{1},'Number')) id=C{2}; end if (strcmp(C{1},'Name')) name=C{2}; end if (strcmp(C{1},'Country')) country=C{2}; end if (strcmp(C{1},'Lat')) lat=C{2}; lat=str2num(lat{1}); end if (strcmp(C{1},'Long')) long=C{2}; long=str2num(long{1}); end if (strcmp(C{1},'Height')) height=C{2}; height=str2num(height{1}); end if (strcmp(C{1},'Start year')) startyear=C{2}; end if (strcmp(C{1},'End year')) stopyear=C{2}; end if (strcmp(C{1},'First Good year')) goodyear=C{2}; end if (strcmp(C{1},'Source ID')) sourceid=C{2}; end if (strcmp(C{1},'Source file')) sourcefile=C{2}; end if (strcmp(C{1},'Normals source')) normalsource=C{2}; end if (strcmp(C{1},'Normals source start year')) normalstart=C{2}; end if (strcmp(C{1},'Normals source end year')) normalend=C{2}; end if (strcmp(C{1},'Normals')) normals=C{2}; nromals=str2num(normals{1}); end if (strcmp(C{1},'Standard deviations source')) stdsource=C{2}; end if (strcmp(C{1},'Standard deviations source start year')) stdstart=C{2}; end if (strcmp(C{1},'Standard deviations source end year')) stdend=C{2}; end if (strcmp(C{1},'Standard deviations')) stds=C{2}; stds=str2num(stds{1}); end end data=textscan(fid,'%d%f%f%f%f%f%f%f%f%f%f%f%f'); % Vector of years for this station year=double(data{1}); % Matrix of temp data bdata=[data{2} data{3} data{4} data{5} data{6} data{7} data{8} data{9} data{10} data{11} data{12} data{13}]; % Vector of temp data sdata=bdata'; sdata=sdata(:); % Vectors of months and years smonth=repmat(1:12,1,length(year)); syear=[]; for k=1:length(year); syear=[syear ones(1,12)*year(k)]; end N=length(sdata); % %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%VISUALISATIONS % Comment/uncomment as needed % Remove nonexistent data sdata(find(sdata<-98))=NaN; % plot full data serially %plot(syear+(smonth-1)/12, sdata) % plot year cycle %plot(sdata(1:(N-3)),sdata(4:N)) % % Plot colored year cycle % clf % hold on % for ii=1:(N-4) % co=ii*[1 1 1]/N; % plot([sdata(ii) sdata(ii+1)],[sdata(ii+3) sdata(ii+4)],'LineWidth',2,'Color',co); % end % set(gca,'Color',[153 136 85]/255) % xlabel('T(t)') % ylabel('T(t+3)') % V=axis; % title(strcat(char(name),' / ',char(country)),'FontSize',20) % % make map %if (height<-100) height=0; end; % remove missing heights %plot3(long,lat,height,'.') %text(long,lat,height,strcat(name,' ',country)); % plot monthly deviation from mean temperature %mm=mean(sdata(find(~isnan(sdata)))); %plot(syear+(smonth-1)/12, sdata-mm); % Plot a polar diagram of temperatures clf hold on singleplot=0; % single color or multicolor mi=-50; ma=50; % Draw background lines theta=2*pi*(0:100)/100; for r=-30:10:30 R=0.1+0.9*(r - mi)/(ma-mi); if (R>0) plot(R.*cos(theta),R.*sin(theta),'k','Color',[0.7 0.7 0.7]) text(R+.025,0+.025,num2str(r),'FontSize',8) end end monthname=['Jan';'Feb';'Mar';'Apr';'May';'Jun';'Jul';'Aug';'Sep';'Oct';'Nov';'Dec']; for mo=1:12 tm=-2*pi*(mo-1)/12; plot([0 Rm*cos(tm)],[0 Rm*sin(tm)],'k','Color',[0.7 0.7 0.7]); text(1.05*Rm*cos(tm),1.05*Rm*sin(tm),monthname(mo,:),'HorizontalAlignment','center'); end theta=-(2*pi*(smonth-1)/12)'; R=0.1+0.9*(sdata - mi)/(ma-mi); Rm=1; if (singleplot==1) plot(R.*cos(theta),R.*sin(theta)) else for ii=1:(N-1) co=ii*[1 1 1]/N; plot([R(ii)*cos(theta(ii)) R(ii+1)*cos(theta(ii+1))],[R(ii)*sin(theta(ii)) R(ii+1)*sin(theta(ii+1))],'Color',co,'LineWidth',2); end end axis([-1.1 1.1 -1.1 1.1]) axis square axis off title(strcat(char(name),' / ',char(country)),'FontSize',20) % Make unreliable mean year plot % bdata(find(bdata<-98))=NaN; % plot(year, mean(bdata')) drawnow pause %%%%%%%%%%%%%%%%%%%%%%%%% % Collect data across stations llong=[llong long]; llat=[llat lat]; hheight=[hheight height]; fclose(fid); end end end end </code>