`rmmissing`

Removes missing entries (NaN, empty text, undefined) from arrays and tables.

Category: Language Fundamentals > Data Types

Syntax

B = rmmissing(A)

Removes elements of a vector, or the rows of a matrix or table, that contain missing values.

B = rmmissing(A, dim)

Removes along dimension dim; 2 removes columns.

B = rmmissing(___, 'MinNumMissing', n)

Removes only rows or columns that have at least n missing values.

[B, TF] = rmmissing(___)

Also returns a logical index of what was removed.

Description

What counts as missing depends on the type: NaN for numbers, <missing> for strings, empty character vectors in a cell array of text, <undefined> for categorical and NaT for datetime. An empty string "" in a string array is not missing.

By default one missing value is enough to remove a whole row (a column when dim is 2). Raise 'MinNumMissing' to require more.

Input arguments

A: Numeric array, string array, cell array of character vectors, categorical or table.
dim: Dimension to remove along.

Output arguments

B: Array without the missing entries.
TF: Logical column or row vector marking the removed rows or columns.

Example

rmmissing([1 NaN 3 4])
rmmissing([1 NaN; 3 4])
rmmissing([1 NaN; 3 4], 2)
T = table([1;NaN;3], ["a";"b";missing], 'VariableNames', {'A','B'});
[R, TF] = rmmissing(T)

Output

ans = 1×3

     1     3     4
ans = 1×2

     3     4
ans = 2×1

     1
     3
R = 1×2 table

    A     B 
    _    ___

    1    "a"
TF = 3×1 logical array

   0
   1
   1

rmmissing