- * TestTarget in your example is a column vector. It must be a row vector in the line above which is why it's transposed.
- * Column 'n' of the 'result' matrix corresponds to TestTarget(n).
Determining if two column adjacent values in a matrix cross over any values in a vector
1 Ansicht (letzte 30 Tage)
Ältere Kommentare anzeigen
Terence Ng
am 3 Jun. 2019
Kommentiert: Adam Danz
am 3 Jun. 2019
Hi there,
I have a matrix (attached as ExampleMatrix.mat), where I want to find where two adjacent column values "cross over" values of a separate vector.
Simpler example with only two columns:
TestMatrix = [...
0,0.533459574621213;
18.5334595746212,18;
36,36.5334595746212;
54,54.5334595746212;
72,72.5334595746212;
90,90.5334595746212;
108,108.533459574621;
126,126.533459574621;
144,144.533459574621;
162,162.533459574621;
180,180.533459574621;
198,198.533459574621;
216,216.533459574621;
234,234.533459574621;
252,252.533459574621;
270,270.533459574621;
288,288.533459574621;
306,306.533459574621;
324,324.533459574621;
342,342.533459574621...
];
TestTarget = [18.25; 306.25];
So far I've been looping over pairs of columns over the values of TestTarget like so:
% Find the difference between two adjacent columns
DifferenceVector = diff(TestMatrix, 1, 2);
% Find + differences
pV_ = sign(DifferenceVector) == 1;
% Find - differences
nV_ = sign(DifferenceVector) == -1;
for i = size(TestTarget, 1);
% For + differences, see if target value is crossed over
passingFromLeft(pV_) = TestMatrix(pV_, 1) + DifferenceVector(pV_) > TestTarget(i);
passingFromRight(pV_) = TestMatrix(pV_, 2) - DifferenceVector(pV_) <= TestTarget(i);
% For - differences, see if target value is crossed over
passingFromLeft(nV_) = TestMatrix(nV_, 1) + DifferenceVector(nV_) < TestTarget(i);
passingFromRight(nV_) = TestMatrix(nV_, 2) - DifferenceVector(nV_) >= TestTarget(i);
passingFromLeft & passingFromRight
end
ans =
20×1 logical array
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
ans =
20×1 logical array
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
Is there a more efficient method to do this without looping over the values of TestTarget, for multiple columns?
0 Kommentare
Akzeptierte Antwort
Adam Danz
am 3 Jun. 2019
Bearbeitet: Adam Danz
am 3 Jun. 2019
After sorting the matrix so that minimum values of each row are on the left, you just need one line to find the intervals that 'skip over' the targets.
% Sort matrix so min val is always on left
TMsort = sort(TestMatrix,2);
result = (TMsort(:,1) < TestTarget.') & (TMsort(:,2) > TestTarget.');
Results:
result =
20×2 logical array
0 0
1 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 0
0 1
0 0
0 0
However, the data you attached in the mat file contains 101 columns so I'm not sure what needs adapted from this 2-column solution. In other words, are you just conserned with the first and last column (assuming they are sorted)?
2 Kommentare
Adam Danz
am 3 Jun. 2019
OK, I couldn't understand if there was a follow-up question or not. If I understand you correctly, you'll need to apply my method to the relevant portion of your matrix by removing it from the irrelevant columns. After you sort by row, if you're just interested in whether the target is skipped over within that row, you just need to look at the first and last (sorted) column within each row.
Let me know if can help more on this matter.
Weitere Antworten (0)
Siehe auch
Kategorien
Mehr zu Shifting and Sorting Matrices finden Sie in Help Center und File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!