How to show only the same variable

Hello,

I have a .mat file as following

Name adress company

BOB london BIM

Alfred Paris BOB

John BOB CEF

I would like to display only the duplicate variable in or order to create this new .mat file

Name adress company

BOB

BOB

BOB

If someone have an idea to create the adapted code?

Thanks in advance

[Chunru]

% Data
varnames = ["Name"    "adress"    "company"];
data = ["BOB"       "london"    "BIM"
        "Alfred"    "Paris"     "BOB"
        "John"      "BOB"       "CEF"];
T = array2table(data, 'VariableNames', varnames);

% Find the duplicate one
Tc = categorical(T{:,:});       % convert from string to categorical
Tcv = Tc(:);                     % make it into a long vector
dup = mode(Tcv);                 % duplicated entries (using mode)

% The location of duplicate one
idx = Tc == dup;
% Generate output
Tout = strings(size(data));
Tout(idx) = dup;
Tout = array2table(Tout, 'VariableNames', varnames)

Tout = 3×3 table
    Name     adress    company
    _____    ______    _______

    "BOB"    ""         ""    
    ""       ""         "BOB" 
    ""       "BOB"      ""    

[Ali]

Thanks a lot,

I thought that I could get my solution using that kind of code

uc1 = unique( Data(:,:));

and this one errase only the duplicate in column not in row visibely

have a nice day.

[Ali]

Sorry For the late answer,

the duplication of yhe entries doesn' t work

dup = mode(Tcv); % duplicated entries (using mode)

It the same variables are randomly put in the table

I needed to find where they are...

[Image Analyst]

You accepted Walter's answer, so we assume everything is working perfectly now.

[Ali]

I accepted walter's code

But chunru's code is more adapted to what I was looking for

walter's code is more complicated to adapt for the moment

so I focused on chunru's code and i saw that the duplication of the entries didn't work

dup = mode(Tcv); % duplicated entries (using mode)

this function doesn't seem to work

the same names (not variables) are randomly put in the table

I should keep the names which are the same and that appear in more than one variable.

Thanks for your understanding

Answer 1

[Walter Roberson]

common_values = intersect(intersect(Name, adress), company);

N = length(Name);
NName = strings(N, 1);
mask = ismember(Name, common_values);
NName(mask) = Name(mask);

Nadress = strings(N, 1);
mask = ismember(adress, common_values);
Nadress(mask) = adress(mask);

Ncompany = strings(N, 1);
mask = ismember(company, common_values);
Ncompany(mask) = company(mask);

output = table(Nname, Nadress, Ncompany, 'VariableNames', {'Name', 'adress', 'company'});

[Ali]

Thank you!

[Ali]

Its perfect for few variables but its very complicated for 50 variables...

[Walter Roberson]

Do you have a table() object? The code can be automated for a table object.

Question: what should be done for a name that appears only in most of the variables but not all of them? Should you keep only names that appear in all variables? Should you keep names that appear in more than one variable?

[Ali]

>>I should keep names which are the same that appear in more than one variable.

>>I don't know if its a table object unfortunately, It's a table which is read like that:

A = readtable('DataFilter.xlsx','TextType','string');

and where I exctract the Variables like that

VarNames = T.Properties.VariableNames;

when I look at the properties:

A.Properties

ans =

struct with fields:

Description: ''

UserData: []

DimensionNames: {'Row' 'Variables'}

VariableNames: {1×46 cell}

VariableDescriptions: {}

VariableUnits: {}

RowNames: {}

I hope it helps.

Thank you for your times!!

[Walter Roberson]

Untested as I do not have a sample of your data.

A = readtable('DataFilter.xlsx','TextType','string');
names_by_var = varfun(@unique, A, 'OutputFormat', 'cell');
[G, ID] = findgroups({names_by_var{:}});
counts = accumarray(G, 1);
names_with_dups = ID(counts>1);
is_dup = varfun(@(V) ismember(V, names_with_dups), A, 'OutputFormat', 'uniform');
Aarray = table2array(A);
B = strings(size(Aarray));
B(is_dup) = Aarray(is_dup);

[Ali]

I have an error when I use it

Error using findgroups (line 77)

A grouping variable must be a categorical, numeric, logical, datetime,

or duration vector, or a cell array of character vectors.

Error in [G, ID] = findgroups({names_by_var{:}});

The code should work with a table as described below

Name adress company

BOB london BIM

Alfred Paris BOB

John BOB CEF

and it should return

Name adress company

BOB

BOB

BOB

[Walter Roberson]

A = readtable('DataFilter.xlsx','TextType','string');
names_by_var = varfun(@unique, A, 'OutputFormat', 'cell');
[G, ID] = findgroups(categorical({names_by_var{:}}));
counts = accumarray(G, 1);
names_with_dups = ID(counts>1);
is_dup = varfun(@(V) ismember(categorical(V), names_with_dups), A, 'OutputFormat', 'uniform');
Aarray = table2array(A);
B = strings(size(Aarray));
B(is_dup) = Aarray(is_dup);

[Ali]

Error using categorical could not find unique values in DATA using the UNIQUE function.

Error in [G, ID] = findgroups(categorical({names_by_var{:}}));

Caused by:

Error using cell/unique Cell array input must be a cell array of character vectors.