How can I seperate these number from a cell and put it into a matrix.

I have tried systax "cell2mat" but it didnt work.

[Star Strider]

You will have to provide (attach) that file if you want help with it. I was able to find it online although not able to download it.

[Hai Luu]

Sorry, here are my files

Answer 1

[Star Strider]

Try this:

[num,txt,raw] = xlsread('2019_nC0v_20200121_20200126_cleaned.xlsx');
vars = regexp(txt(1,:), ',', 'split');
varsc = string(vars{:})
newcell = regexp(txt(2:end), ',', 'split');

C1 = cellfun(@(x)str2double(x(:,1)), newcell);
C2 = cellfun(@(x)x(:,2), newcell);
C3 = cellfun(@(x)x(:,3), newcell);
C4 = datetime(cellfun(@(x)x(:,4), newcell), 'InputFormat','MM/dd/yyyy');
for k = 1:4
    C(:,k) = cellfun(@(x)str2double(x(:,k+4)), newcell);
end

COVID19 = table(C1,C2,C3,C4,C(:,1),C(:,2),C(:,3),C(:,4));
COVID19.Properties.VariableNames = {'Nr','Province/State','Country','Date Last Updated','Confirmed','Suspected','Reecovered','Deaths'};

and:

Check = COVID19(1:5,:)

produces:

Check = 

    Nr    Province/State         Country          Date Last Updated    Confirmed    Suspected    Reecovered    Deaths
    __    ______________    __________________    _________________    _________    _________    __________    ______

    0      {'Shanghai'}     {'Mainland China'}       21-Jan-2020           9           10            0           0   
    1      {'Yunnan'  }     {'Mainland China'}       21-Jan-2020           1            0            0           0   
    2      {'Beijing' }     {'Mainland China'}       21-Jan-2020          10            0            0           0   
    3      {'Taiwan'  }     {'Taiwan'        }       21-Jan-2020           1            0            0           0   
    4      {'Jilin'   }     {'Mainland China'}       21-Jan-2020           0            1            0           0   

I cannot get the variable names to import correctly or parse correctly. I added them manually.

[Hai Luu]

Thanks in advance, I have to say that without your code I cannot do this.

[Hai Luu]

I still have this question is instead of using for loop

for k = 1:4
    C(:,k)=cellfun(@(x)str2doublex(x(:,1:k)),newcell);
end

Can I use

C = cellfun(@(x)str2double(x(:,5:8)),newcell);

I have tried and it appeared error:

Error using cellfun

Non-scalar in Uniform output, at index 1, output 1.

Set 'UniformOutput' to false.

[Star Strider]

I tried that first.

To use that approach, do this:

Cc = cellfun(@(x)str2double(x(:,5:8)),newcell, 'Uni',0);
C = cell2mat(Cc);

The rest of the code is unchanged, with these two assignments replacing the loop. I am not certain which is more efficient.

[Hai Luu]

Cool. Any possible way to classify the people come from Main Land, Taiwan,.... for instance.?

[Star Strider]

I am not certain what you mean by ‘classify’. If you want to assign numbers to them, use the findgroups function.

For example —

CountryID = findgroups(COVID19(:,3));

You can do the same for 'Province/State' by referring to column 2 instead.

To combine them, use either of these:

CountryStateID = findgroups(COVID19(:,[2 3]));
CountryStateID = findgroups(COVID19(:,2), COVID19(:,3));

Both produce the same result.

There are other options as well. See: Grouping and Binning Data in the Preprocessing Data documentation section.

Answer 2

[Jon]

You will not be able to put your data from the cell array you show to a matrix because you have both numeric and non-numeric data in your cell array and a MATLAB matrix can be either numeric or non-numeric but not both.

I suggest instead that you store your data in a table where it will be very natural to work with it

[Hai Luu]

the reason I want to put it into a matrix is because I want to do something with machine learning :D. I have tried put numbers like this file but with systax cell2mat, I still cannot put it into a matrix.

[Star Strider]

@Hai Luu — Please see my Answer!

↓