Description
intersect(a,b) returns a row vector of unique values
present in both a and b arrays. Values are
sorted in increasing order
- for complex numbers : par increasing magnitudes, then by increasing phases
- for text : in alphabetical order.
If a and b are tables or timeseries, then intersect
returns the rows present in both a and b. In the case of timeseries, the row times are taken into account unlike
the row names for tables.
 | Two NaN elements are always considered as different. So NaN or rows or columns having
NaN will never be in the result M. |
[M, ka, kb] = intersect(a,b) additionally returns the vectors
ka and kb of indices in a
and b of selected components firstly met, such that
M=a(ka) and M=b(kb).
In the case of tables and timeseries, M = a(ka,:) and M = b(kb, :).
Common rows or columns
When the orient argument is provided, the comparison is performed
between the rows of a and b -- each one being
considered as a whole --, or between their columns.
intersect(a,b,"r") or intersect(a,b,1) will return
the matrix of stacked unduplicated rows met in both a and
b, sorted in lexicographic ascending order.
If a and b don't have the same number of columns,
[] is returned without comparing the values.
[M,ka,kb] = intersect(a,b,"r") additionally returns the vectors
ka and kb of the minimal indices of common rows,
respectively in a and b,
such that M=a(ka,:) and M=b(kb,:).
intersect(a,b,"c") or intersect(a,b,2) does
the same for columns.
| This option is not supported for table and timeseries. |
Examples
A = grand(3, 3, "uin", 0, 9)
B = grand(2, 4, "uin", 0, 9)
intersect(A, B)
[N, ka, kb] = intersect(A,B);
ka, kb |  |  | |
--> A = grand(3, 3, "uin", 0, 9)
A =
0. 6. 4.
6. 6. 6.
2. 7. 9.
--> B = grand(2, 4, "uin", 0, 9)
B =
1. 8. 0. 2.
6. 2. 2. 1.
--> intersect(A, B)
ans =
0. 2. 6.
--> [N, ka, kb] = intersect(A,B);
--> ka, kb
ka =
1. 3. 2.
kb =
5. 4. 2.
In the above example, note that 6 is met four times in A, at indices [2 4 5 8].
Only the minimal index 2 is returned in ka. Same situation for 2 in B.
NaN values can never be in the result:
%nan == %nan
intersect([1 -2 %nan 3 6], [%nan 1:3]) | | | |
--> %nan == %nan
ans =
F
--> intersect([1 -2 %nan 3 6], [%nan 1:3])
ans =
1. 3.
intersect() can also process some characters or some texts. Since Scilab is great with UTF
characters, here is an example with some Arabic contents, getting characters present in
both sentences:
A = strsplit("هو برنامج علمي كبير ""Scilab""")'
B = strsplit("فهو حر ومفتوح")'
intersect(A,B) | | | |
--> A = strsplit("هو برنامج علمي كبير ""Scilab""")'
A =
!ه و ب ر ن ا م ج ع ل م ي ك ب ي ر " S c i l a b " !
--> B = strsplit("فهو حر ومفتوح")'
B =
!ف ه و ح ر و م ف ت و ح !
--> intersect(A,B)
ans =
! ر م ه و !
Column-wise or Row-wise processing of two matrices: Here we process 2 matrices of
signed 1-byte integers, and get the common columns:
A = int8(grand(3,5,"uin",0,1))
B = int8(grand(3,9,"uin",0,1))
[M,ka,kb] = intersect(A, B, "c");
M, ka, kb | | | |
--> A = int8(grand(3,5,"uin",0,1))
A =
0 0 1 1 1
0 0 1 1 0
0 0 0 0 1
--> B = int8(grand(3,9,"uin",0,1))
B =
1 0 1 1 1 0 1 1 1
1 0 0 1 1 1 0 0 0
1 0 1 0 1 1 1 0 0
--> [M,ka,kb] = intersect(A, B, "c");
--> M, ka, kb
M =
0 1 1
0 0 1
0 1 0
ka =
1. 5. 3.
kb =
2. 3. 4.
intersect() for booleans is mainly useful with the "r" or "c" option.
Here is an example with a sparse boolean matrix:
[F, T] = (%f, %t);
A = [F F T F T F ; T F F T T T ; T T F T F F]
B = [F T F T F F ; T F F F T F ; F T F F T F]
[M,ka,kb] = intersect(A, sparse(B), "c");
issparse(M), full(M), ka, kb | | | |
--> A = [F F T F T F ; T F F T T T ; T T F T F F]
A =
F F T F T F
T F F T T T
T T F T F F
--> B = [F T F T F F ; T F F F T F ; F T F F T F]
B =
F T F T F F
T F F F T F
F T F F T F
--> [M,ka,kb] = intersect(A, sparse(B), "c");
--> issparse(M), full(M), ka, kb
ans =
T
ans =
F F T
T T F
F T F
ka =
6. 1. 3.
kb =
1. 5. 4.
intersect() for duration and datetime:
A = hours([1 7 7 4 3 6 6]);
B = hours([7 5 3 0]);
[M, ka, kb] = intersect(A, B)
A = datetime("today") + A;
B = datetime("today") + B;
[M, ka, kb] = intersect(A, B) | | | |
intersect() for table:
A = table([1 7 7 4 3 6 6]', [1:7]');
B = table([7 5 3 0]', [2;2;5;5]);
[M, ka, kb] = intersect(A, B)
A = table([1;3;4;2], ["d";"c";"f";"h"], [%f;%t;%t; %f], "VariableNames", ["double", "string", "boolean"]);
B = table([2;3;4;1], [%t;%t;%t;%t], ["d";"c";"f";"h"], "VariableNames", ["double", "boolean", "string"]);
[M, ka, kb] = intersect(A, B) | | | |
intersect() for timeseries: