How to read text strings and count them

Hello All,

Lets assume I have one file which contains strings. I want to read that file, and create a matrix of word count/line. For example, I define I will have 5 words in file, which are

"we, are, you, people, student"

and the text file contains following three lines.

we are people and we are students you are students and people

Now I want matrix

        we, are,    you,    people,   student
line 1   2   2       0       1          1
line 2   0   1       1        1          1

[Adam]

doc textscan
doc strtok
doc strcmp
doc ismember

should help you (one or other of strcmp or ismember rather than both), though I don't have time to give a more detailed answer so just a comment. Someone else may provide a more detailed answer if you cannot work out the right combination of those functions.

Answer 1

[Guillaume]

One fairly simple way:

searchwords = {'we', 'are', 'you', 'people', 'student'}; %list of words to search
filecontent = fileread('somefile.txt'); %read whole text file
filelines = regexp(filecontent, '\r?\n', 'split'); %split lines
filewords = regexp(filelines, '[^a-zA-Z]+', 'split'); %split each line into words
wordcount = zeros(numel(filelines), numel(searchwords));
for line = 1:numel(filelines)
  wordcount(line, :) = cellfun(@(searchword) sum(strcmp(filewords{line}, searchword)), searchwords);
end

[Aravin]

Perfectory what I wanted. In addition, could you please tell how could I create searchwords by input file. For example, from file each word should be obtained and than their count matrix should be calculated.

Now I will have only file somefile.txt and the search words should be

{'and', 'are', 'you', 'people', 'student', 'we'}

[Guillaume]

To get the the unique list of words in the file, after it's been split into words:

searchwords = unique([filewords{:}]);

Note that I've edited my answer to improve the regular expression that splits the lines into words (it now ignores consecutive non-letters).