How to convert the contents of a cell array into specific numbers?

Hi guys, I wrote this code to convert the columns 1 and 2 of my cell array into specific numbers as mentioned in the code but no value changed in column 1 and no error is shown even!.

Maybe there is a problem of indexing anyone can help me out plz?

   [row_patho, col_patho]=size(Pro_patho_data);
   for i=1:1:row_patho
   if Pro_patho_data{i,1}>=70
     Pro_patho_data{i,1}=11;
   end
   end
   for k=1:1:row_patho
   if strcmpi(Pro_patho_data{k,2},'t3b')
     Pro_patho_data{k,2}=3;
  end
  end

Inputs:

73 t2b

59 t1c

58 t3b

78 t3b

wanted_outputs:

11 t2b

59 t1c

58 3

11 3

[Walter Roberson]

I suspect that the entries in column 1 are character vectors, not numeric

[chocho]

@Walter Roberson no, entries in column1 are all numeric, except some entries are in the form of '6.40E+01' ...

Answer 1

[Guillaume]

Pro_patho_data = {73, 't2b'; 59, 't1c'; 58, 't3b'; 78, 't3b'}; %demo data

col1 = cell2mat(Pro_patho_data(:, 1));
col1(col1 > 70) = 11;
Pro_patho_data(:, 1) = num2cell(col1);

Pro_patho_data(strcmpi(Pro_patho_data(:, 2), 't3b'), 2) = {3}

[chocho]

@Guillaume, Thank you for your quick reply! I tried your code and got this error

Error using cat Dimensions of matrices being concatenated are not consistent.

Error in cell2mat (line 83) m{n} = cat(1,c{:,n});

Error in Pathological_data_cleaning (line 28) col1 = cell2mat(Pro_patho_data(:, 1));

[Guillaume]

One of the row in your first column contains more (or less) than one number. What should happen in this case?

With your original code, if all numbers in the matrix in that cell were greater than 70, the matrix would have been replaced by just one number, your 11. If any of the numbers in the matrix was below 70 then no number would have been changed.

[chocho]

@Guillaume friend, i have shared my data can you have a look at it

[Stephen23]

Why not use textscan to import the first column as numeric, and the second as char?

[chocho]

@Stephen Cobeldick how plz help me on this issue

[Guillaume]

"i have shared my data can you have a look at it"

Now, we're on a completely different problem. Clearly, you haven't imported that text file properly in matlab if you get an error with my code. I've no idea how you imported it, but the simplest way to import it is with:

pro_patho_data = readtable('Pro_patho_data.txt')

pro_patho_data will then be a table, which is actually easier to manipulate than a cell array. Possibly you'll want to give better names to the columns of the table but that's up to you.

If the data is stored as a table, then it's even easier to do your replacements:

pro_patho_data.Var1(pro_patho_data.Var1 > 70) = 11;
pro_patho_data.Var2(strcmp(pro_patho_data.Var2, 't3b')) = {3}

However, I wouldn't recommend replacing a char array by a number as you then end up with a column that contains both text and numbers which is going to be more difficult to process. Possibly, you'll want {'3'} instead of {3} so the whole column is still text.

[chocho]

@Guillaume I used textscan to import my data

fid= fopen('Patho_clean_data2.txt','rt');
format = repmat('%q',[1 7 ]);
Pro_patho_data = textscan(fid,format,'Delimiter','\t');
fclose(fid);
Pro_patho_data=[Pro_patho_data{:}];
[row,col]=size(Pro_patho_data);

Var1 and Var2 are Unrecognized variable name.

[chocho]

this is the whole data i used

[Walter Roberson]

%q means to import quote-delimited strings. You would need '%f%s%f%s%f%f%f'

[chocho]

@Walter Roberson, I have updated the format and i tried my code again but nothing changed

Answer 2

[Birdman]

Assuming that Inputs is a cell array:

 inputs={'73','t2b';'59','t1c';'58','t3b';'78','t3b'}
 wanted_outputs=regexprep(inputs,{'\<7\w*','\<t3\w*'},{'11','3'})

[chocho]

@Birdman this code convert all the columns in my data which fall in these cases but i want to convert every column separately based on some specific conditions which i have to specify, can you plz make it more specific column by column like my code. with my respect to you

[Birdman]

Use a for loop:

inputs={'73','t2b';'59','t1c';'58','t3b';'78','t3b'};
expr={'\<7\w*','\<t3\w*';'11','3'};
outputs=num2cell(zeros(size(inputs)));
for i=1:size(inputs,2)
    outputs(:,i)=regexprep(inputs(:,i),expr(1,i),expr(2,i));
end

[chocho]

 @Birdman friend, I got this error with your code and plz I shared my data can you have a look?

Index exceeds matrix dimensions.

Error in Pathological_data_cleaning (line 31) outputs(:,i)=regexprep(Pro_patho_data(:,i),expr(1,i),expr(2,i))

[chocho]

index exceeds matrix dimensions.

Error in Patho_data_grouping (line 13) outputs(:,i)=regexprep(inputs(:,i),expr(1,i),expr(2,i));