I want to read this text file from Matlab directly and want to import data of all columns into work space in an array.

fid = fopen('1.txt');
a = textscan(fid,'%s%s%s%s%s%s%s','Delimiter','','Headerlines',4);
lines=a{1:7};
fclose(fid)

Answer 1

[Walter Roberson]

Your file appears to contain 10 lines. You ask to skip the first 4, which would leave 6. You then ask to read 7 lines. You are going to have difficulty with the 7th value.

[Haqqah Khursheed]

fid = fopen('1.txt');
a = textscan(fid,'%s%s%s%s%s%s%s','Delimiter','','Headerlines',3);
lines=a{1:7};
fclose(fid)

i want to arrange string and numeric data into separate arrays. My actual file contains 44640 rows and 7 columns.

[Walter Roberson]

fid = fopen('1.txt', 'rt');
datacell = textscan(fid, '%s%s%f%f%f%f%f', 'HeaderLines',3);
fclose(fid);
timestamps = strjoin(datacell{1}, {' '}, datacell{2});
numerics = cell2mat(datacell(3:end));

[Walter Roberson]

You can then convert the timestamps to datetime objects with

datetime(timestamps, 'Format', 'yy-MM-dd HH:mm')

Answer 2

[Jeremy Hughes]

I suggest the 'CollectOutput' parameter

fid = fopen('1.txt');
a = textscan(fid, '%s%s%s%s%s%s%s', 'Delimiter', '', 'Headerlines', 4, 'CollectOutput', true);
lines = a{1};
fclose(fid)

Also, if you're trying to read just lines, this will work:

fid = fopen('1.txt');
a = textscan(fid, '%[^\r\n]', 'EndOfLine', '\r\n', 'Headerlines', 4, 'Delimiter', '','Whitespace','');
lines = a{1};
fclose(fid);

Your first code might work for the file you have, but the second block is a more robust version for line reading if you have leading spaces, or if you want to capture empty lines.

Answer 3

[Wei-Rong Chen]

It will be much easier if you use 'readtable' function. For example:

dataTable = readtable('1.txt','Delimiter',your_delimiter,'ReadVariableNames',set_true_if_first_row_is_VarNames);
num_array = table2array(dataTable(:,1:7));

[Walter Roberson]

In R2016b, readtable will interpret the date string '15-04-01' as a date 0015-04-01 unless you specify 'datetimetype', 'text'.

It is possible to specify a format for readtable for the date handling

readtable('1.txt', 'format', '%{yy-MM-dd}D%{HH:mm}D%f%f%f%f%f')

Unfortunately for this purpose it does not seem to be possible to grab the date and time as a single entity, so you end up with two datetime objects in the output and you cannot just add them together.

[Walter Roberson]

If you use

t = readtable('1.txt', 'format', '%{yy-MM-dd}D%{HH:mm}D%f%f%f%f%f');

then

timestamps = t.Var2 - dateshift(t.Var2, 'begin', 'day') + t.Var1;

This takes into account several arcane behaviours of datetime objects.