MATLAB Answers

0

Import data from txt file

Latest activity Commented on by Adam Danz
on 16 Jul 2019
Hi there,
my txt file looks like this: (see attached txt file)
What I try to get is a simple table with lines like this:
frame; timestamp [s]; loc_x; loc_y; loc_z
1007; 52045.728; -135.966; 225.786; -52.285
1008; 52045.745; -135.966; 225.786; -52.285
How can I do this? Thanks

  3 Comments

Adam Danz
on 16 Jul 2019
I see the frame and timestamp components. Where are the xloc yloc etc.?
What are the numbers 1007; 52045.728; -135.966; 225.786; -52.285?
frame 1010 ts 52045.778 nbod 4 nfly 0 nmea 0 nmearef 0 nhand 0 nmar 0 nhuman 0 ninertial 4
bod 0 not tracked
bod 1 not tracked
bod 2 qu 1.000 loc -135.967 225.786 -52.283 rot -0.282 -0.960 0.001 0.960 -0.282 -0.008 0.008 -0.001 1.000
bod 3 not tracked
no human model data
inertial body 0 st 0 error 0.000
inertial body 1 st 0 error 0.000
inertial body 2 st 2 error 0.000
loc -135.967 225.786 -52.283 rot -0.282 -0.960 0.001 0.960 -0.282 -0.008 0.008 -0.001 1.000
inertial body 3 st 0 error 0.000
Rik
on 16 Jul 2019
Do you need code that only works on this file, or should it also handle multiple bodies being tracked?
What have you tried so far on your own?
In line 1 of each block is the frame number (1007) and the timestamp (52045.728 seconds)
In line 10 of each block is the location and rotation data. Starting with location (loc_x, loc_y, loc_z) then rotation (rot_xx, rot_xy, rot_xz; rot_yx, rot_yy, rot_yz; rot_zx, rot_zy, rot_zz).
All other lines of each block are not necessary.

Sign in to comment.

2 Answers

Adam Danz
Answer by Adam Danz
on 16 Jul 2019
Edited by Adam Danz
on 16 Jul 2019
 Accepted Answer

This uses readtable() to read in your txt file and it uses regexp() with the 'tokens' flag to pull out relevant data.
%% Import data into table
opts = delimitedTextImportOptions("NumVariables", 1);
opts.DataLines = [1, Inf];
opts.VariableNames = "txt";
test1 = readtable("test1.txt", opts); %file is on path
%% Extract desired data
% Pull out the frame and ts data
tokens = regexp(test1.txt, 'frame (\d+) ts (\d+.\d+)', 'tokens');
frame_ts = cell2mat(cellfun(@str2double,[tokens{:}]','UniformOutput',false));
% Pull out loc coordinates
tokens = regexp(test1.txt, '^loc (-?\d+.\d+) (-?\d+.\d+) (-?\d+.\d+)', 'tokens');
loc = cell2mat(cellfun(@str2double,[tokens{:}]','UniformOutput',false));
% Check that frame_ts and loc have same number of rows.
if size(frame_ts,1) ~= size(loc,1)
error('''frame_ts'' and ''loc'' have different number of rows.')
end
%% Put data into table
T = array2table([frame_ts, loc], 'VariableNames', {'frame', 'timestamp', 'loc_x', 'loc_y', 'loc_z'});
% inspect first few rows
head(T)
Result
format long
>> head(T)
ans =
8×5 table
frame timestamp loc_x loc_y loc_z
_____ _________ ________ _______ _______
1007 52045.728 -135.966 225.786 -52.285
1008 52045.745 -135.966 225.786 -52.285
1009 52045.761 -135.967 225.786 -52.283
1010 52045.778 -135.967 225.786 -52.283
1011 52045.795 -135.967 225.786 -52.283
1012 52045.811 -135.967 225.786 -52.283
1013 52045.828 -135.967 225.786 -52.283
1014 52045.845 -135.967 225.786 -52.283

  0 Comments

Sign in to comment.


Answer by Ferdinand Grosse-Dunker on 16 Jul 2019

This works great! Thanks a lot Adam!

  1 Comment

Adam Danz
on 16 Jul 2019
Glad I could help!

Sign in to comment.