paddata

Pad data by adding elements

Since R2023b

Syntax

B = paddata(A,m)

B = paddata(A,m,Name=Value)

Description

B = paddata(A,m) pads A to size m by adding elements to the trailing side of A. For example, for a scalar size m:

• If A is a vector, then paddata(A,m) pads A to length m.

• If A is a matrix, table, or timetable, then paddata(A,m) pads A to have m rows.

• If A is a multidimensional array, then paddata(A,m) pads A to the size specified by m along the first dimension whose size is greater than 1.

If m is less than or equal to the size of A in the operating dimension, then paddata returns all of A.

B = paddata(A,m,Name=Value) specifies additional parameters for padding using one or more name-value arguments. For example, paddata(A,m,Pattern="circular") pads by repeating the input data circularly.

Examples

Pad Vector

Create a four-element column vector, and pad the vector to six elements. By default, the paddata function adds zeros to the trailing size of the numeric vector.

A = [1; 3; 5; 7];
B = paddata(A,6)

B = 6×1

     1
     3
     5
     7
     0
     0

Pad a row vector to six elements.

A2 = [2 4 6 8];
B2 = paddata(A2,6)

B2 = 1×6

     2     4     6     8     0     0

Match Length of Another Vector

Create two vectors with different lengths.

A1 = [2; 8; 3; 5];
A2 = [9; 4; 6; 2; 7; 7; 0];

Determine the length of the longer vector.

szA1 = size(A1,1);
szA2 = size(A2,1);
[m,idx] = max([szA1 szA2])

m = 7

idx = 2

Pad the shorter vector to match the length of the longer vector.

B1 = paddata(A1,m)

B1 = 7×1

     2
     8
     3
     5
     0
     0
     0

You can concatenate vectors of the same length. Create a matrix using the two vectors.

C = [B1 A2]

C = 7×2

     2     9
     8     4
     3     6
     5     2
     0     7
     0     7
     0     0

Pad Matrix

Create a 3-by-3 matrix. Pad the columns to a length of 4 by adding one element to each column. Pad the rows to a length of 6 by adding three elements to each row.

A = [1 3 5; 2 4 6; 7 8 10]

A = 3×3

     1     3     5
     2     4     6
     7     8    10

B = paddata(A,[4 6])

B = 4×6

     1     3     5     0     0     0
     2     4     6     0     0     0
     7     8    10     0     0     0
     0     0     0     0     0     0

Add Array Page

Create a 3-by-3 matrix as the first page in a 3-D array. Add a second page to the array by padding along the third dimension.

A = [1 3 5; 2 4 6; 7 8 10]

A = 3×3

     1     3     5
     2     4     6
     7     8    10

B = paddata(A,2,Dimension=3)

B = 
B(:,:,1) =

     1     3     5
     2     4     6
     7     8    10


B(:,:,2) =

     0     0     0
     0     0     0
     0     0     0

Specify Fill Values for Table Variables

Create a timetable with variables of different data types.

num = [10; 20; 30];
cat = categorical(["A"; "B"; "A"]);
log = logical([1; 0; 1]);
str = ["Text 1"; "Text 2"; "Text 3"];
TT = timetable(num,cat,log,str,Timestep=hours(2))

TT=3×4 timetable
    Time    num    cat     log       str   
    ____    ___    ___    _____    ________

    0 hr    10      A     true     "Text 1"
    2 hr    20      B     false    "Text 2"
    4 hr    30      A     true     "Text 3"

Pad each timetable variable. paddata uses the default fill value for the data type of each variable. The default fill value for each data type is the same as the value of elements that MATLAB® creates when assigning a value beyond the last row of the table.

B1 = paddata(TT,6)

B1=6×4 timetable
    Time     num        cat         log        str   
    _____    ___    ___________    _____    _________

    0 hr     10     A              true     "Text 1" 
    2 hr     20     B              false    "Text 2" 
    4 hr     30     A              true     "Text 3" 
    6 hr      0     <undefined>    false    <missing>
    8 hr      0     <undefined>    false    <missing>
    10 hr     0     <undefined>    false    <missing>

Specify a custom fill value for the elements to add to each timetable variable. paddata extends the row times, so you do not need to specify a fill value for the row times.

B2 = paddata(TT,6,FillValue={mean(num),"C",1,""})

B2=6×4 timetable
    Time     num    cat     log       str   
    _____    ___    ___    _____    ________

    0 hr     10      A     true     "Text 1"
    2 hr     20      B     false    "Text 2"
    4 hr     30      A     true     "Text 3"
    6 hr     20      C     true     ""      
    8 hr     20      C     true     ""      
    10 hr    20      C     true     ""      

Repeat First Element

Create a column vector, and pad the vector by repeating the leading element.

A = [1; 3; 5; 7];
B = paddata(A,7,Pattern="edge",Side="leading")

B = 7×1

     1
     1
     1
     1
     3
     5
     7

Input Arguments

A — Input data

vector | matrix | multidimensional array | table | timetable | cell array | structure array

Input data, specified as a vector, matrix, multidimensional array, table, timetable, cell array, or structure array.

Note

If A is a cell array, then paddata changes the size of the entire array. It does not change the size of each cell in the array. Use the cellfun function to apply paddata to each cell in a cell array.

m — Size of padded data along operating dimension

nonnegative integer scalar | vector of nonnegative integers

Size of padded data along operating dimension, specified as a nonnegative integer scalar or vector of nonnegative integers. Each element represents the size of the padded data in an operating dimension.

• If m is greater than the size of A in the operating dimension, then paddata adds elements.

• If m is less than or equal to the size of A in the operating dimension, then paddata returns the input data without adding elements.

Name-Value Arguments

Specify optional pairs of arguments as Name1=Value1,...,NameN=ValueN, where Name is the argument name and Value is the corresponding value. Name-value arguments must appear after other arguments, but the order of the pairs does not matter.

Example: B = paddata(A,m,Pattern="circular")

Dimension — Dimensions to operate along

"auto" (default) | positive integer scalar | vector of positive integers

Dimensions to operate along, specified as "auto", a positive integer scalar, or a vector of positive integers. Each element represents a dimension of the input data.

If Dimension is "auto", then the operating dimension depends on the input arguments:

• If m is a scalar and A is an array, then the operating dimension is the first dimension whose size is greater than 1.

• If m is a vector, then the operating dimensions are 1:numel(m).

• If A is a table or timetable, then the operating dimension is 1, and operation is along each table or timetable variable separately.

FillValue — Fill value

[] (default) | scalar | cell array | scalar structure

Fill value for added elements, specified as a scalar, cell array, or scalar structure.

The default fill value for each class is the same as the value of elements that MATLAB^® creates when assigning a value past the end of a vector. For example, the default fill value for numeric input data is 0.

• If A is an array, then a scalar FillValue indicates the value for all elements added to A during padding.

• If A is a table or timetable, then a cell array FillValue indicates a different fill value for elements added to each table or timetable variable. The number of cells in the cell array must match the number of table or timetable variables. If A is a table with row names, then paddata extends the row names with the default row name, such as RowN; it does not pad the row names using FillValue. If A is a timetable with row times, then paddata extends the row times; it does not pad the row times using FillValue.

• If A is a structure array, then a scalar structure FillValue indicates a different fill value for each field in the input data. The number and names of fields in FillValue must match the number and names of fields in the input data.

If you specify FillValue, you cannot specify Pattern.

Pattern — Pattern for adding elements

"constant" (default) | "edge" | "circular" | "flip" | "reflect"

Pattern for adding elements, specified as one of the pattern names in the table. The pattern is repeated until the resized data has size m.

• If A is a table with row names, then paddata extends the row names with the default row name, such as RowN; it does not add to the row names using Pattern.

• If A is a timetable with row times, then paddata extends the row times; it does not add to the row times using Pattern.

If you specify Pattern, you cannot specify FillValue.

This table lists the pattern names with a description and a sample of how each pattern pads the input data A = [1 2 3].

Pattern Name	Description	Padded Data
"constant"	Pad data with the default value determined by the data type of A.
"edge"	Pad data by replicating the leading and trailing endpoints as constant fill values.
"circular"	Pad data by repeating the input data circularly.
"flip"	Pad data by flipping the input data symmetrically. Endpoints are duplicated.
"reflect"	Pad data by reflecting the input data. Endpoints are not duplicated.

Side — Side of input data for padding

"trailing" (default) | "leading" | "both"

Side of input data for padding, specified as one of these values:

• "trailing" — Pad A with trailing elements.

• "leading" — Pad A with leading elements.

• "both" — Pad A on both sides. If the number of elements to add in the operating dimension is even, then pad the trailing and leading sides of A evenly. If the number of elements to add in the operating dimension is odd, then pad the remaining element on the trailing side of A.

Tips

• paddata only adds elements to the input data. paddata is recommended if you do not want to remove elements from your data and you do not require the resized data to match the target size. If you require the resized data to respect the sizes in m, then consider using the resize function, which can also remove elements from your data.

Extended Capabilities

GPU Arrays

Accelerate code by running on a graphics processing unit (GPU) using Parallel Computing Toolbox™. (since R2024a)

This function fully supports GPU arrays. For more information, see Run MATLAB Functions on a GPU (Parallel Computing Toolbox).

Distributed Arrays

Partition large arrays across the combined memory of your cluster using Parallel Computing Toolbox™. (since R2024a)

This function fully supports distributed arrays. For more information, see Run MATLAB Functions with Distributed Arrays (Parallel Computing Toolbox).

Version History

Introduced in R2023b