Plotting array filtered by column?

How would I go about graphing columns filtered by specific results in a third column?

The following table as an example, how would I plot only the second and third column entries that are 1.0000 in the first?

A =

    1.0000    1.0000    5.5000
    1.0000    2.0000    1.2000
         0    3.0000    6.4000
    1.0000    4.0000    3.7000
         0    5.0000    3.6000
         0    6.0000    9.6000

[Dyuman Joshi]

Try comparing with the first column values with 1.000, ideally with a tolerance, and use logical indexing to get the corresponding values in 2nd and 3rd column.

Answer 1

[Jon]

So to make @Dyuman Joshi suggestion more concrete

% Define example points
A = [
    1.0000    1.0000    5.5000
    1.0000    2.0000    1.2000
         0    3.0000    6.4000
    1.0000    4.0000    3.7000
         0    5.0000    3.6000
         0    6.0000    9.6000];

% Define tolerance for finding rows to plot
tol = 1e-6;

% plot rows meeting critiera that element in first column is a 1
idl = ismembertol(A(:,1),1,tol); % or alternatively idl = abs(A(:,1)-1)<=tol
plot(A(idl,2),A(idl,3))

[Jon]

Did this answer your question?

[Brandon]

Would this method also work if I'm trying to filter it by a string?

[Jon]

If you are filtering by a string then I assume the match would be exact, no tolerance involved. Also you couldn't store the numerical data and the strings in a double array, you would have to use some other structure, like a cell array. Here are some ways you could filter by a string:

% Define example points
A = {
    'ABC'    1.0000    5.5000
    'ABC'    2.0000    1.2000
    'DEF'    3.0000    6.4000
    'ABC'    4.0000    3.7000
    'XYZ'    5.0000    3.6000
    'PQT'    6.0000    9.6000};


% plot rows meeting critiera that element in first column is 'ABC'
idl = strcmp(A(:,1),'ABC');
figure
plot([A{idl,2}],[A{idl,3}]); % use [] to converts comma separated list to vector

% Or maybe a clearer way to do this, is to use a table
T = cell2table(A); % columns are named by default as  A1,A2,A3
idl = strcmp(T.A1,'ABC');
figure
plot(T.A2(idl),T.A3(idl))

[Brandon]

Quick followup, say I wanted to filter the first column by elements that are either 1 or 2 using the idx and {:,1} method, what would that look like?

[Jon]

If you only want to plot rows where the element in the first column is either a 1 or a 2, then there are no tolerances involved. For this you can use basic logical indexing.

% Define example points
A = [
    4    1.0000    5.5000
    1    2.0000    1.2000
    2    3.0000    6.4000
    3    4.0000    3.7000
    0    5.0000    3.6000
    2    6.0000    9.6000];

% Plot rows meeting critiera that element in first column is a 1 or a 2
idl = A(:,1)==1 | A(:,1)==2

idl = 6×1 logical array
   0
   1
   1
   0
   0
   1

plot(A(idl,2),A(idl,3),'o-')

[Brandon]

Got it, thanks. I asked because I had tried that but using || instead of | since I'm used to java haha

[Dyuman Joshi]

You can also use ismember for that.

[Jon]

Yes, that's cleaner and connects back to the beginning of the the thread where the match was approximate and we had to use ismembertol.

% Define example points
A = [
    4    1.0000    5.5000
    1    2.0000    1.2000
    2    3.0000    6.4000
    3    4.0000    3.7000
    0    5.0000    3.6000
    2    6.0000    9.6000];

% Plot rows meeting critiera that element in first column is a 1 or a 2
idl = ismember(A(:,1),[1,2])

idl = 6×1 logical array
   0
   1
   1
   0
   0
   1

plot(A(idl,2),A(idl,3),'o-')