Ploting data from .txt file

I gathered some data in this format

17:51:10 927.00
17:51:15 950.00
17:51:21 827.00
17:51:27 999.00
17:51:33 1088.00
17:51:38 1094.00
17:51:44 1109.00

as you can see first part is time and second is the data.

I extracted data using this code

fid = fopen('.\test.txt','r');
cac = textscan(fid, '%f%f%f%f','Delimiter', ':');
data=cac{:,4};
HH=cac{:,1};
MM=cac{:,2};
SS=cac{:,3};

And tried to plot them using a timeseries function

ts1 = timeseries(data,(HH:MM:SS));

To my dissapointment it doesn't plot and the ts1 variable seems to be invalid.

Is there some other way of ploting time data ? Or am I using the timeseries function wrong ?

[Slawomir Kania]

BTW, all the data seem to produce correct horizontal arrays data, HH - hours, MM - minutes, SS - seconds

[Von Duesenberg]

What if you simplify your workflow, load your data with the readtable function, convert your table to a timetable with the table2timetable function, and plot the first column against the second?

Answer 1

[dpb]

timeseries doesn't plot anything; all it does is create a timeseries object. Per the documentation, the data for the time vector is either an array of doubles or date strings; you passed neither format but a cell array.

Try

c=textscan(fid,'%s%f');         % read time as cell, data as float
ts=timeseries(c{2},c{1});       % build a timeseries object therefrom
fid=fclose(fid);

plot(ts)

I've not used the timeseries object enough to know how to clean up the display into something more user-friendly, but that gets the base object created for you to play with further.

Frankly, what little I've tried to 'spearmint with it, the timeseries just seems to be in the way more than help as compared to just using datetime or the venerable datenum

Mayhaps someone with more specific experience can elucidate further than what does the documentation on how to convert the above to more pleasing units in seconds and all...