How to create a new group of character array in matlab? - arrays

Question

I have data stored in text files.The data is in 'cell array of string' after read it using textscan and contains various of colour name. Below is the content of my data:
name of colour
'lavender'
'lavenderblush'
'lemonchiffon'
'lightblue'
'lightcoral'
'lightcyan'
I want to create new array to group all color characters into the main color only (red, blue, orange, brown,etc).
I am really struggling to solve this problem. Thank you in advance for any help.
load_data = fopen('result.txt', 'r');
C = textscan(load_data, ' %s ');
fclose(load_data);
name = C{1,1};
group = char(name)
if group{:,1} == lavender
fprintf('purple');
else
fprintf('nothing');
end
This is my code but if I run this, always get error
Cell contents reference from a non-cell array object.

Answer 1

From your comment above, I assume you're having trouble with your code, not actually with the sorting of colors, or how to categorize them.
group is not a Cell array, it's a Char array. So in order to access its values you should use group(:,1) instead of group{:,1}
Remember that the number of columns in group is the number of characters in that line, normalized to the number of characters of the largest string in that set. So 2 issues here:
You can't use group(:,1), as it will get the first character of all the strings in that array. You should get the entire line for that string group(1,:). NB: I say string for simplicity, it's actually a char array.
'lavender' has only 10 characters, but it will have 15, as per the largest string. So the string comparison doesn't quite work, unless you add the extra blank spaces to compensate
You can try out the code below:
load_data = fopen('result.txt', 'r');
C = textscan(load_data, ' %s ');
fclose(load_data);
name = C{1,1};
if name{1,:} == '''lavender'''
fprintf('purple');
else
fprintf('nothing');
end
I assume that your TXT file actually has the string 'lavender', in this case I used character escape '''lavender'''.

Answer 2

The error you are getting is fairly clear. Once you have called char(name), you are working with a regular character array (i.e. a string) and no longer a cell array. Braces (i.e. {}) are used to index cell arrays so you should instead be using parentheses (i.e. ()). Note this actually happens as soon as you index using {} so C{1,1} would actually return a string already and the char line is thus redundant:
if group(:,1) == lavender
I suspect you actually want something more like
name = C{1};
group = char(name) %this line is redundant because C{1} already extracts a string
if strcmp(group,lavender)
fprintf('purple');
else
fprintf('nothing');
end
but it's impossible to say since you did not define the variable lavender.
I would also question what data structure you intend to use. I'm going to assume you actually have some way of categorizing the colours? I'm going to assume this is manual in which case I would suggest converting your txt file to a csv file and putting your manual colour categories as a second column but I'll leave the implementation details to you.
Lets say you have 3 colour categories for now, 'purple', 'blue', and 'orange', my suggestion is to use a logical matrix that has 3 columns (1 per colour category) and n rows where n is the number of rows in your text file (i.e. the number of colours you need to categorize).
Now I'm going to assume you have some sort of mapping function that can categorize your colours so map('lavender') returns 'purple' and map('lightcyan') returns 'blue'
First we should make a cell array of categories that we can use to map the category string to its column number:
categories = {'purple'
'blue'
'orange'}
and the result will go in the logical matrix categorized
load_data = fopen('result.txt', 'r');
C = textscan(load_data, ' %s ');
fclose(load_data);
n = numel(C);
categories = {'purple'
'blue'
'orange'};
categorized = false(n, numel(categories)); %preallocation
for row = 1:n
colour = C{row};
category = map(colour); %you need to implement this map function yourself.
categorized(row, strcmp(category, categories)) = true;
end