# counting frequency of letters in text using matlab

suppose we have following text

``````s='i love georgia and its nature';
``````

what i want is to count frequency of occurrence of each letter(space are not included of course) and sketch some chart(for example bar chart), first i have created code which counts letters using map container

`````` function character_count(s)
% s is given string and given program will count  occurence of letters in
% sentence
MAP=containers.Map();% initialize   MAP for frequency counting
n=length(s); % get length of given string
letters=unique_without_space_sorting(s);
for ii=1:n
if ~isletter(s(ii))==1
continue;
elseif  isKey(MAP,s(ii) )
MAP(s(ii)) = MAP(s(ii))  + 1;
else
MAP(s(ii)) = 1;

end
end
y=values(MAP);
y= cell2mat(y);
bar(y);
set(gca,'xticklabel',letters)

end
``````

here function

``````letters=unique_without_space_sorting(s);
``````

returns cell array of letters of string s without sorting and space, here is its corresponding code

`````` function cell_stirng=unique_without_space_sorting(s)
s=regexprep(s,'[^\w'']','');
[~, idxs, ~] = unique(s, 'last');
s= s(sort(idxs));
n=length(s);
cell_stirng=cell(n,1);
for jj=1:n
cell_string{jj}=s(jj);
end
end
``````

when i run this code, i got following image

as you see there is no labels on x axis, how can i fix this issue? thanks in advance

You can use the first output of `unique` to give you the unique values and use those as your x labels

``````[values, idxs, ~] = unique(s, 'last');

% Make sure that they aren't sorted
[~, sortind] = sort(idxs);
values = num2cell(values(sortind));

% And later after creating your bar plot
set(gca, 'xtick', 1:numel(values), 'XTickLabels', values);
``````

Or rather than doing that, you can just use the `stable` input to `unique` to make sure that they are in the order of appearance instead.

``````S = lower(strrep(s, ' ', ''));
[values, ~, b] = unique(S, 'stable');

hist(b, unique(b))
set(gca, 'xtick', 1:numel(values), 'xticklabels', num2cell(values))
``````

Or if you want a histogram of all letters

``````S = lower(strrep(s, ' ', ''));
counts = histcounts(double(S), double('a':'z'));
bar(counts)
set(gca, 'xtick', 1:26, 'xticklabels', num2cell('a':'z'))
``````

``````str = 'i love georgia and its nature';

num_times = zeros(26,1);
letters = {'a', 'b', 'c', 'd', 'e', 'f', 'g', 'h', 'i', 'j', 'k', 'l', 'm', ...
'n', 'o', 'p', 'q', 'r', 's', 't', 'u', 'v', 'w', 'x', 'y', 'z'};

for k = 1:length(str)

% Convert letter to its lower case, get ASCII value,
% a = 97, b = 98, ..., z = 122
n = uint16(lower(str(k)));

% If character between 'a' and 'z'
if n < 122 && n > 97

% Convert to be between 1 and 26
n = n - 96;

% Index count array with n
num_times(n) = num_times(n) + 1;

end

end

clf
stem(num_times);
set(gca, 'XTick', 1:26);
set(gca,'XTickLabel', letters)
``````

Output:

If you don't want to change your other code, see the last 2 lines of mine for labelling x-axis.

Edit:

You could use these lines for plotting instead of the above, so that only letters which have a non-zero frequency are plotted

``````clf
stem(num_times(num_times ~= 0));
set(gca, 'XTick', 1:sum(num_times ~= 0));
set(gca,'XTickLabel', letters(num_times ~= 0))
``````

Output: