delif

Purpose

Deletes rows from a matrix. The rows deleted are those for which there is a 1 in the corresponding row of e.

Format

new_x = delif(x, e)
Parameters:
  • x (NxK matrix) – data

  • e (Nx1 logical vector) – vector of 0’s and 1’s

Returns:

new_x (MxK matrix) – consisting of the rows of x for which there is a 0 in the corresponding row of e. If no rows remain, delif() will return a scalar missing.

Examples

Basic usage with column vector

// Create column vector
x = { 1.5,
      0.8,
      0.7,
      1.2,
      1.9,
      0.2,
      2.0 };

// Create logical vector of 1's and 0's
e = x .> 1;

/*
** Assign 'new_x' to be equal to 'x'
** with the rows removed in which 'e' equals 1
*/
new_x = delif(x, e);

After the code above, new_x should equal:

0.8
0.7
0.2

Matrix case

In this example, we will remove all observations in which the value of the third column is 3.

// Create a matrix with 3 columns
x = { 20    10     2,
      33    13     3,
      37    12     2,
      34    12     3,
      35     8     1,
      25    15     2,
      34     8     2,
      37     8     1,
      37     3     1,
      31     4     1 };

// Create logical vector of 1's and 0's
e = x[., 3] .== 3;

/*
** Assign 'new_x' to be equal to 'x' without
** the rows in which the third column equals 3
*/
new_x = delif(x, e);

After the code above, new_x should be equal to:

20     10      2
37     12      2
35      8      1
25     15      2
34      8      2
37      8      1
37      3      1
31      4      1

Create new ‘x’ and ‘y’ based on ‘y’

In this example, we will remove all observations from x and y in which the value of the third column of y is 2.

// Create 'y' matrix
y = { 1,
      1,
      0,
      2,
      0,
      1,
      1,
      0,
      0,
      2 };

// Create 'x'  matrix
x = { 1.6841  -0.1203,
     -1.0433   0.2564,
      1.2207  -1.4388,
      0.7423   0.2133,
      0.7288   1.0434,
      0.8115   1.8166,
     -0.3230   1.4763,
      1.2944   0.7635,
      1.3839   0.6648,
     -0.6330   0.4845 };


// Create logical vector of 1's and 0's
e = y .== 2;

/*
** Assign 'new_x' to be equal to 'x' without
** the rows in which 'y' equals 2
*/
new_x = delif(x, e);

// Remove all observations in which 'y' equals 2
new_y = delif(y, e);

After the code above, new_y and new_x should equal:

new_y = 1   new_x = 1.6841  -0.1203
        1           -1.0433   0.2564
        0           1.2207  -1.4388
        0           0.7288   1.0434
        1           0.8115   1.8166
        1           -0.3230   1.4763
        0           1.2944   0.7635
        0           1.3839   0.6648

Logical comparison of multiple columns

x = { 0 10 20,
     30 40 50,
     60 70 80 };

// Logical vector, comparing two columns
e =(x[., 1] .gt 0) .and (x[., 3] .lt 100);

new_x = delif(x, e);

After the code above:

new_x = 0 10 20

Remarks

The input e will usually be generated by a logical expression using dot operators. For instance:

/*
** Create a vector 'e' with a 1 for each row in which the
** value in the second column of 'x' is less than 100,
** otherwise a 0
*/
e = x[., 2] .> 100;

new_x = delif(x, e);

Or the equivalent statement:

new_x = delif(x, x[., 2] .> 100);

will delete all rows of x whose second element is greater than 100. The remaining rows of x will be assigned to y.

See also

Functions selif()