Rewrite first 3 columns of csv without re-writing the entire file

I have large csv files that need the first 3 columns rewritten. Is it possible to do that without rewritting the rest of the columns to save computing time. I will need to skipp the first two rows which are header strings and the first column that I am rewritting is a time/date string while the next two columns are numbers.

Here is what I have so far

    FileList = dir('*.csv');

       % get the file name:
       filename = FileList(1).name

    %%%%%%%%%Renames files according to date %%%%%%%%%%%
    presg=filename(1:11);
    mid='CORR';
    sufsg=filename(12:end);
    sufsg2=filename(12:26);

    lf=strcat(presg,sufsg2);
    load(lf);

    % IIBOP_128HzCORR_2013_07_18_000.csv

    newName=strcat(presg,mid,sufsg);
    zipName=strcat(presg,mid,sufsg2);
    movefile(filename,newName);

    %%Display File Start Date/Time and End Date/Time
    tcnt=length(DATA(:,1)); % total cnts HS***********

    fstdate=datestr(DATA(1,1), 'mm/dd/yyyy HH:MM:SS:FFF')
    fenddate=datestr(DATA(tcnt,1), 'mm/dd/yyyy HH:MM:SS:FFF')

    %%Enter offsets (s)
    rn1=0;
    rn2=0;
    rn3=0;

    cp1=datenum('07/14/2013 20:40:02:445', 'mm/dd/yyyy HH:MM:SS:FFF');
    os1=rn1*(1/(24*3600));
    cp2=datenum('07/14/2013 20:40:03:203', 'mm/dd/yyyy HH:MM:SS:FFF');
    os2=rn2*(1/(24*3600));
    cp3=datenum('07/14/2013 20:40:03:203', 'mm/dd/yyyy HH:MM:SS:FFF');
    os3=rn3*(1/(24*3600));

    DATA2=zeros(tcnt,1:3);
    DATA2(1:tcnt,2)=DATA(1:tcnt,2);

    for i=1:tcnt;
        if DATA(i,1)<=cp1;
            DATA2(i,1)=DATA(i,1);
            DATA2(i,3)=DATA(i,3);
        else if DATA(i,1)>cp1 && DATA(i,1)<=cp2;
                DATA2(i,1)=DATA(i,1)+os1;
                DATA2(i,3)=DATA(i,3)+rn1;
            else if DATA(i,1)>cp2 && DATA(i,1)<=cp3;
                    DATA2(i,1)=DATA(i,1)+os1+os2;
                    DATA2(i,3)=DATA(i,3)+rn1+rn2;
                else
                    DATA2(i,1)=DATA(i,1)+os1+os2+os3;
                    DATA2(i,3)=DATA(i,3)+rn1+rn2+rn3;
                end
            end
        end
    end

    samp=tcnt;

    csvwrite(newName,DATA2,2,0);

    zip(zipName,newName);

The original file was writen like this:

% Adate_string=datestr(xt(1:samp,1),'mm/dd/yyyy HH:MM:SS:FFF');
% Cdate=cellstr(Adate_string);
% 
% Headers = {'Datetime','TimerTick','Relatvie_Time',...
%     'String_Torque','String_Weight','String_Pressure',...
%     'String_Acceleration_Z','String_Rotational_Velocity',...
%     'StringSense_Heart_Beat','StringSense_Voltage',...
%     'StringSense_Temperature','Node_RSSI','Base_RSSI'};
% 
% FUnits = {'Datetime','Unitless','Seconds',...
%     'Kft-Lbs','KIP','Psi',...
%     'G','RPM',...
%     'Unitless','DC_Voltage',...
%     'C','dBm','dBm'};

% fid = fopen(newName, 'W');  % Uppercase, no text mode
% fprintf(fid, '%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s\r\n', ...
%         Headers{:}, FUnits{:});
% 
% 
% for i=1:samp;
%      fprintf(fid, '%s,', Cdate{i,1});
% for cc = 1:12
%     if cc<12
%   fprintf(fid, '%.10f,', Data_EX(i, cc));
%     else
%         fprintf(fid, '%.10f\r\n', Data_EX(i, cc));
%     end
% end
% end
% fclose(fid);

Here is the file formatt:

Datetime  TimerTick  Relatvie_Time  String_Torque  String_Weight  String_Pressure  String_Acceleration_Z  String_Rotational_Velocity  StringSense_Heart_Beat  StringSense_Voltage  StringSense_Temperature  Node_RSSI  Base_RSSI
Datetime  Unitless  Seconds  Kft-Lbs  KIP  Psi  G  RPM  Unitless  DC_Voltage  C  dBm  dBm
07/14/2013 20:40:02:445  16187  126.4609375  0.043643963  0.50814116  -76.4645462  0.062795378  1.104599714  0.998359978  6.136816025  37.9807663  -44  -51
07/14/2013 20:40:02:453  16188  126.46875  0.043643963  0.50814116  -71.68567657  0.05592538  1.104599714  0.998359978  6.136816025  37.9807663  -44  -49

Thank You,

Ian

Answer 1

[Walter Roberson]

In the case where you are changing a string to another string of exactly the same size, it is possible to rewrite a file in place. This will require extra computation time and I/O operations compared to creating a new file with the changed data. You would not gain any benefit with lines that short.