Organizing and Accessing Data in MATLAB

This reference shows common use cases, so it is not an exhaustive list.

Representing Data
Homogenous
Data Type	Icon	Purpose	Syntax
Double, single, (u)int8, (u)int16, (u)int32, (u) int64, complex		Numeric arrays, matrix computations, math	`[1,2,3], [1;2;3], uint8(), int16()`
String		Text arrays	`"hello world"`
Char		Single characters, character arrays	`'hello'`
Categorical		Discrete, nonnumeric data	`categorical()`
Datetime		Absolute dates and timestamps, including time zones	`datetime('July 12, 2001 08:15:01')`
Duration		Elapsed times	`duration(h,m,s), hours(), minutes()`
Calendar of duration		Relative time based on calendar	`caldays(), calweeks()`
Logical		True/false, test state, identify data by condition	`logical(), ==, ~=, >, >=, <, <=, &, &&, \|, \|\|`
Other specialized types		sparse, enumeration, custom, …	» Documentation
Heterogeneous
Data Type	Icon	Purpose	Syntax
Table		Mixed-type, column-oriented data (spreadsheet-like). Store metadata	`table(x,y,z), array2table`
Timetable		Timestamped tabular data	`timetable(t,x,y) table2timetable,array2timetable`
Structure		Fields can contain data of any size and type. Ideal for nonrectangular data.	`struct()`
Cell array		Each cell in the array can contain any data type, any size	`cell(), {pi,ones(5), "hello"}`
Tall array		MATLAB data types can be made "tall" when data does not fit in memory	`ds = datastore(), T = tall(ds)`
Dictionary		Object that maps unique keys to values	`d = dictionary(keys,values)`

Data Selection

Use array indexing to select data.

Linear indexing for 1D arrays:
X(1) First element
X(end) Last element

Row, column indexing for multidimensional arrays:
A(1,2)
A(1,1,2)

Select multiple with a vector:
A([1,3],1)

Use colon : to select a range:
A(1:3,1)
A(:,1) All rows, column 1
A(1,:) Row 1, all columns
A(1:2:end,:) Every other row

Remove data from array:
A(1,:) = [];

Logical Indexing
Use logical expressions to select data Elements of X greater than 7: `Y = X(X > 7)` Combine conditions `X(X > 3 & X <= 7)` `X(X > 3 \| X <= 7)` Multidimensional arrays Use condition to identify rows or columns: `idx = X > 7; X(idx, :)` Tables and timetables `T(:,vartype('numeric')) TT(timerange(t1,t2),:)`

Logical Indexing

Use logical expressions to select data

Elements of X greater than 7:
Y = X(X > 7)

Combine conditions
X(X > 3 & X <= 7)
X(X > 3 | X <= 7)

Multidimensional arrays
Use condition to identify rows or columns:
idx = X > 7; X(idx, :)

Tables and timetables
T(:,vartype('numeric')) TT(timerange(t1,t2),:)

Container Indexing

Using parentheses () for indexing retains the initial data type. Access the underlying data with curly braces {}. Tables and structures also allow you to reference data by name.

» Examples

Type	Subset	Contents
Table	Returns a table: `T(1,2) T(:,"A") T(:,["A","B"])`	Returns underlying data: `T{1,2} T{:,"A"} T.A` `T.RowsT.Variables`
Timetable	Same as above: `TT("Apr 1,... 2004",5)`	Same as above: `TT.Time`
Cell array	Returns a cell: `C(1,2)`	`C{1,2} C{:} -> comma separated list`
Structure	Returns a struct: `S(1,1)`	`S.Field`