CODEWITHSUNDEEP
matlabfile-io
Alex Knok
Alex Knok

Reputation: 19

Read and plot .txt file (into .mat)

I need to read test.TXT file(tab delimited) into MATLAB.

TXT file have form:

Datum Time Data1 Data2 Data3 Data4 Data5
06/28/2012 09:27,3 1,931764 -0,008698 4,151306 33,865424 -44,923096
06/28/2012 09:27,3 1,931764 -0,003662 4,154358 33,865424 -44,831543
06/28/2012 09:27,4 1,928712 -0,001526 4,168701 33,866102 -45,472412
.....................................................

After that I need to plot these data (for example time vs. data1; time vs. data2..)

What is the easiest way to open and read test.txt file so I can then plot my data (insert all data to workspace..). I tried to import data, but then MATLAB import data as one column. :/

Thanks

EDIT

I just noticed that I copy wrong format of time into first post :S

Datum Time Data1 Data2 Data3 Data4 Data5
06/28/2012 17:09:27,3 1,931764 -0,008698 4,151306 33,865424 -44,923096
06/28/2012 17:09:27,3 1,931764 -0,003662 4,154358 33,865424 -44,831543
06/28/2012 17:09:27,4 1,928712 -0,001526 4,168701 33,866102 -45,472412
.....................................................

17:09:27,4 -> 17(hour), 09 (minutes), 27 (seconds), 4(miliseconds)

Here is now big problem with , and .

Upvotes: 0

Views: 7884

Answers (1)

Salain
Salain

Reputation: 752

As answered previously in this question you should use either textscan or fscanf

After taking a closer look there are a few interesting nuances to your question. The solution I went with was to preserve the columnar structure by reading everything as a string. That helps to keep everything ordered while commas are replaced with periods, date strings are processed to datenums and finally the numbers stored as strings are converted.

fid = fopen('test.txt','rt');
header = textscan(fid,'%s %s %s %s %s %s %s',1); 
data = textscan(fid,'%s %s %s %s %s %s %s');
data = cellfun(@(x) strrep(x,',','.'),data,'UniformOutput',false);
clean_data(:,1) = arrayfun(@(x) datenum([data{1}{x} ' ' data{2}{x}]), 1:length(data{1}) )';
clean_data(:,2:6) = cell2mat(cellfun(@str2num , [data{3:end}],'UniformOutput',false));
fclose(fid);

There are a few messy transitions in there. But the general approach should be solid so long as your files are not outrageously long. Here are the results:

>> datestr(clean_data(:,1))
ans =
28-Jun-2012 09:27:18
28-Jun-2012 09:27:18
28-Jun-2012 09:27:24

>> clean_data(:,2:end)
ans =
1.9318   -0.0087    4.1513   33.8654  -44.9231
1.9318   -0.0037    4.1544   33.8654  -44.8315
1.9287   -0.0015    4.1687   33.8661  -45.4724

And finally to plot versus date, which was the original goal:

>> plot(clean_data(:,1),clean_data(:,2))
>> datetick

Upvotes: 2

Related Questions