Reformat data

1 vue (au cours des 30 derniers jours)
Rahul
Rahul le 29 Mar 2012
I have the following data: each set is as follows
code jobnum date time type status
0 1831 3.15000000000000 17:59:0 3 submitted
1 1831 3.15000000000000 17:59:2 1 exectuing
5 1831 3.15000000000000 18:38:49 0 terminated
0 1832
1 1832
5 1832
how do I get it as
jobnum date submitted executing terminated s_type exec_type
1831 3.15 17:59:0 17:59:2 18:38:49 3 1
1832

Réponse acceptée

Kye Taylor
Kye Taylor le 29 Mar 2012
Assuming that your data lives inside someTextFile.txt, that the values in the data set are delimited/separated by tabs (\t) or spaces and that "exectuing" will actually be "executing", the following script formats the data like you showed, then writes this data to newTextFile.txt.
fid = fopen('someTextFile.txt');
inMatlab = textscan(fid,'%*f%f%f%s%f%s','headerlines',1);
fclose(fid);
% get unique jobnums
uJobs = unique(inMatlab{1});
newRow = cell(length(uJobs)+1,7);
% create header for new text file
newRow(1,:) = ...
{'jobnum','date','submitted','executing','terminated','s_type','exec_type'};
% create new rows for new text file
for j = 1:length(uJobs)
isJob = inMatlab{1} == uJobs(j); % which rows belong to this job
isSub = strcmp('submitted',inMatlab{5});% which rows related to submission
isExe = strcmp('executing',inMatlab{5});% which rows related to execution
isTer = strcmp('terminated',inMatlab{5});%which rows related to termination
jobDate = unique(inMatlab{2}(isJob));
% make sure jobDate is indeed unique
assert(isscalar(jobDate),'Same job on two different days?');
% create new row
newRow(j+1,:) = {uJobs(j),...
jobDate,...
char(inMatlab{3}(isSub & isJob)),...
char(inMatlab{3}(isExe & isJob)),...
char(inMatlab{3}(isTer & isJob)),...
inMatlab{4}(isSub & isJob),...
inMatlab{4}(isExe & isJob)};
end
% open new file
fid = fopen('newTextFile.txt','wt');
fprintf(fid,'%s\t%s\t%s\t%s\t%s\t%s\t%s\n',newRow{1,:});
for j = 2:size(newRow,1)
fprintf(fid,'%.0f\t%.2f\t%s\t%s\t%s\t%.0f\t%.0f\n',newRow{j,:});
end
fclose(fid);

Plus de réponses (0)

Catégories

En savoir plus sur Programming dans Help Center et File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by