read_prevah_input

 PUPROSE: read input tabular data for PREVAH hydrological model
 -------------------------------------------------------------------
 USAGE: [data, header] = read_prevah_input(file)
 Input:  file - filename and location
         nodata - nodata value (optional, if given value will be replaced
         with NaN)
 Output: R:    x*n column, where n is the number of timesteps and x are
               the number of input columns, according to specification
               in the PREVAH manual, part III)
               YYYY MM DD HH ... ... ...
         header - 4*1 cell: elevation, x and y coordinate, name

 EXAMPLE: [data, header] = read_prevah_input('rain2003.tab')
 
 See also: write_prevah_grid, read_prevat_grid, read_prevah_result

 Felix Hebeler, Geography Dept., University Zurich, Feb 2009

0001 function [R, header] = read_prevah_input(file,nodata)
0002 % PUPROSE: read input tabular data for PREVAH hydrological model
0003 % -------------------------------------------------------------------
0004 % USAGE: [data, header] = read_prevah_input(file)
0005 % Input:  file - filename and location
0006 %         nodata - nodata value (optional, if given value will be replaced
0007 %         with NaN)
0008 % Output: R:    x*n column, where n is the number of timesteps and x are
0009 %               the number of input columns, according to specification
0010 %               in the PREVAH manual, part III)
0011 %               YYYY MM DD HH ... ... ...
0012 %         header - 4*1 cell: elevation, x and y coordinate, name
0013 %
0014 % EXAMPLE: [data, header] = read_prevah_input('rain2003.tab')
0015 %
0016 % See also: write_prevah_grid, read_prevat_grid, read_prevah_result
0017 %
0018 % Felix Hebeler, Geography Dept., University Zurich, Feb 2009
0019 
0020 %
0021 data = nan(1,4);
0022 
0023 %open file
0024 fid = fopen(file,'r');
0025 
0026 % read header
0027 % read first line and ignore
0028 title = fgetl(fid);
0029 % read second line
0030 remain = fgetl(fid);
0031 col=1;
0032 while true  % loop once over elevation to find number of columns
0033    [str, remain] = strtok(remain);
0034     if isempty(str),  break;  end
0035    header{1,col}=str;
0036    col=col+1;
0037 end
0038 col=col-1;
0039 remain = fgetl(fid);
0040 for i=1:col %loop again for xcoords in 3rd row
0041    [str, remain] = strtok(remain);
0042     if isempty(str),  break;  end
0043    header{2,i}=str;
0044 end
0045 remain = fgetl(fid);
0046 for i=1:col %loop 3rd time for ycoords in 3rd row
0047    [str, remain] = strtok(remain);
0048     if isempty(str),  break;  end
0049    header{3,i}=str;
0050 end
0051 remain = fgetl(fid);
0052 for i=1:col %loop 4th time for name in 4th row
0053    [str, remain] = strtok(remain);
0054     if isempty(str),  break;  end
0055    header{4,i}=str;
0056 end
0057 
0058 % construct format string for data
0059 format = '%d %d %d'; % first 3 columns are read in as integers (YY MM DD )
0060 for i=4:col
0061     format = [format,' %n']; % add a %n to read in double for each data column
0062 end
0063 data = textscan(fid,format);
0064 
0065 %intialise result array
0066 l=length(cell2mat(data(1,col))); % get length of last column, this is used because result files can contain gibberish from some point on,
0067 % which will not be read, but the column length (no of timesteps) will not
0068 % be correct anymore for the reformating
0069 R=nan(l,col);
0070 
0071 % reformat to column vector
0072 for i= 1:col
0073     R(:,i)=data{1,i}(1:l); % l (normally, l equals end)
0074 end
0075 clear data;
0076 % check if nodata values should be replaced
0077 if exist('nodata','var')
0078     R(R==nodata)=NaN;
0079 end
0080 %close file
0081 fclose(fid);