operator '==' not supported for operands of type 'cell'

While there are some useful answers to this type of question, a lot of them are too advanced for me to understand/use as a beginning MATLAB student. I am trying to loop through rows of a table to find instances in which in a particular index is equal to something. Here is a picture of what my table looks like:

and I am interested in comparing when a subject has responded true or false to a certain verb ('smashed' is one of 3). My idea for doing this would be to make a for loop such as the following:

smashedFalseCounter = 0;
for i = 1:height(myTable)
    if myTable.verb(i) == 'smashed'
        if expTable2.response(i) == false
            smashedFalseCounter = smashedFalseCounter + 1;
        end
    end
end

And I get the following error message:

[Mohammad Sami]

Use string comparison functions.

strcmp(myTable.verb{i}, 'smashed')

Answer 1

[madhan ravi]

ismember(myTable.verb(i), 'smashed')

[madhan ravi]

And by the way you don’t need a loop.

Counts = nnz(ismember(myTable.verb(myTable.Response == true), 'smashed'))

Answer 2

[Gizachew Dessalegn]

for s = 1:ncntry
    for r = 1:ncntry
        Esr14{s,r} = zeros(nsec, 1);
        if r ~= s
            for t = 1:ncntry
                if t ~= s && t ~= r
                    % Access cell contents using {} and perform operations
                    term = (va_cs(1,t) * Bmat(t,s))';  % Now numeric values
                    Esr14{s,r} = Esr14{s,r} + term .* EXGR_FNL_cs_c(s,r);
                end
            end
        end
    end  
end

Operator '*' is not supported for operands of type 'cell'.

[Walter Roberson]

Based on the above code, we can guess that either va_cs or Bmat are cell arrays. The code might have to be

term = (va_cs{1,t} * Bmat(t,s))';

or

term = (va_cs(1,t) * Bmat{t,s})';

or

term = (va_cs{1,t} * Bmat{t,s})';

There is also a significant chance that EXGR_FNL_cs_c is a cell array, so the line after that might need to be

        Esr14{s,r} = Esr14{s,r} + term .* EXGR_FNL_cs_c{s,r};

Answer 3

[Steven Lord]

Another possibility for the original question would be to create a pivot table or to use groupcounts or other grouping functions. Let's make a sample table:

subject = (1:9).';
verbs = {'smashed', 'crushed'};
n = numel(verbs);
verb = verbs(randi(n, 1, 9)).';
response = rand(9, 1) > 0.5;
t = table(subject, verb, response)

t = 9x3 table
    subject       verb        response
    _______    ___________    ________

       1       {'crushed'}     false  
       2       {'smashed'}     false  
       3       {'smashed'}     false  
       4       {'crushed'}     true   
       5       {'crushed'}     true   
       6       {'smashed'}     true   
       7       {'crushed'}     false  
       8       {'smashed'}     true   
       9       {'crushed'}     true   

Now let's create a pivot table with response as the variables and the verbs as the rows.

P = pivot(t, Columns='response', Rows='verb')

P = 2x3 table
       verb        false    true
    ___________    _____    ____

    {'crushed'}      2       3  
    {'smashed'}      2       2  

Or take the counts of each group defined by unique combinations of verb and response.

G = groupcounts(t, {'verb', 'response'})

G = 4x4 table
       verb        response    GroupCount    Percent
    ___________    ________    __________    _______

    {'crushed'}     false          2         22.222 
    {'crushed'}     true           3         33.333 
    {'smashed'}     false          2         22.222 
    {'smashed'}     true           2         22.222