loc

Syntax

loc(obj, rowFilter, [colFilter], [view=false])

Arguments

obj is a matrix object. It can be a standard matrix, an indexed series or an indexed matrix.

rowFilter/colFilter can be:

  • a Boolean vector - the rows/columns marked as true will be returned. The length of the vector must match the number of rows/columns of obj.

  • a scalar, a vector or a pair whose data type is compatible with the row/column labels of obj. A pair indicates the selection range (both upper bound and lower bound are inclusive).

Note:

  • If rowFilter/colFilter is a pair, then obj must be an indexed series or an indexed matrix.

  • Data type compatibility rules:

    • INT, SHORT, LONG and CHAR are compatible

    • FLOAT and DOUBLE are compatible

    • STRING and SYMBOL are compatible

view is a Boolean value. The default value is false indicating the result will be a copy of the original matrix (deep copy). If set to true, the result will be a view on the original matrix (shallow copy) and changes made to the original matrix will be reflected in the view.

Details

Access a group of rows and columns of a matrix by label(s) or a boolean vector. Return a copy or a view of the original matrix.

Examples

m=rand(12, 3:4)
m;
col1 col2 col3 col4
3 10 6 5
4 11 6 0
7 2 1 8 
a = m.loc(colFilter=[true, true, true, false], view=true)
b = m.loc(colFilter=[true, true, true, false], view=false)
a;
col1 col2 col3
3 10 6
4 11 6
7 2 1 
b;
col1 col2 col3
3 10 6
4 11 6
7 2 1 
// a view reflects changes made to the original matrix whereas a copy doesn't
m[0,0] = -1
a;
col1 col2 col3
-1 10 6
4 11 6
7 2 1 
b;
col1 col2 col3
3 10 6
4 11 6
7 2 1 
m = rand(48, 6:8)
m;
col1 col2 col3 col4 col5 col6 col7 col8
27 31 47 21 12 43 22 11
3 20 13 37 3 46 27 27
13 5 14 11 26 42 4 18
45 9 31 33 12 19 42 17
2 19 30 25 36 27 21 6
9 36 15 10 29 37 31 42 
// filter with Boolean values
m.loc(rowFilter=[true, true, false, false, true, false])
col1 col2 col3 col4 col5 col6 col7 col8
27 31 47 21 12 43 22 11
3 20 13 37 3 46 27 27
2 19 30 25 36 27 21 6 
m.loc(colFilter=[true, true, false, false, true, false, false, true])
col1 col2 col3 col4
27 31 12 11
3 20 3 27
13 5 26 18
45 9 12 17
2 19 36 6
9 36 29 42 
// filter with labels
m.rename!(`A`A`B`A`B`B, 2022.01.01 + 0..7)
m;
label 2022.01.01 2022.01.02 2022.01.03 2022.01.04 2022.01.05 2022.01.06 2022.01.07 2022.01.08
A 27 31 47 21 12 43 22 11
A 3 20 13 37 3 46 27 27
B 13 5 14 11 26 42 4 18
A 45 9 31 33 12 19 42 17
B 2 19 30 25 36 27 21 6
B 9 36 15 10 29 37 31 42 
m.loc(rowFilter=`A);
label 2022.01.01 2022.01.02 2022.01.03 2022.01.04 2022.01.05 2022.01.06 2022.01.07 2022.01.08
A 27 31 47 21 12 43 22 11
A 3 20 13 37 3 46 27 27
A 45 9 31 33 12 19 42 17 
m.loc(colFilter=2022.01.02);
label 2022.01.02
A 31
A 20
B 5
A 9
B 19
B 36 
m.loc(rowFilter=`B, colFilter=2022.01.03)
label 2022.01.03
B 14
B 30
B 15
When rowFilter / colFilter is a pair, obj must be an indexed matrix, which can be converted using setIndexedMatrix!.
m = rand(30, 5:6).rename!(1..5, 2022.01.01 + 0..5)
m.setIndexedMatrix!()
m;
label 2022.01.01 2022.01.02 2022.01.03 2022.01.04 2022.01.05 2022.01.06
1 5 27 26 18 29 3
2 11 12 21 15 3 3
3 1 23 29 17 7 18
4 1 6 12 27 23 23
5 15 7 3 19 4 8 
m.loc(rowFilter=2:4, colFilter=2022.01.03:2022.01.06)
label 2022.01.03 2022.01.04 2022.01.05 2022.01.06
2 21 15 3 3
3 29 17 7 18
4 12 27 23 23

Related function: at