Variable object (Multivariate Analysis Toolbox for Matlab®)

Holds information about values of a measurement of some scalar magnitude (namely, a feature or a coordinate) over samples. Each variable has a range of possible values, which can be numerical or verbal descriptions. Variables are classified by their level, according to the relationships between their possible values. A variable is nominal if there are no relationships between the possible values of the variable, which are typically non-numerical. For example, a wine primary taste can be described as citrus, woody, or fruity. A variable is ordinal if the values have an order. For example, a person can be described as thin, normal and fat. Finally, a variable is numerical if its values are numerical. We do not make a discrimination between numerical values that are on a ratio scale (e.g., a person's height) and those that are not (e.g., temperature in Celsius).

If the values are verbal descriptions, they are represented as integers, and the connection between them and the verbal descriptions are kept in the lookup table lut. The lut field is therefore always defined for nominal variables, always not defined for numerical variables, and sometimes defined for ordinal variables.

If the variable is known to be distributed according to some probability distribution, this distribution is designated in the field distribution. This field is a structure with two fields - the name of the distribution (e.g., 'normal') and its parameters (e.g, mean and variance). Statistical measures of the variable, namely mean, variance and minmax, are also kept. Each of these is a structure with a 'population' field and with 'sample' field, keeping the population and sampling measures, respectively. Unknown population measures are designated by NaNs.

Someimes, a transformation is applied to a variable (e.g., log-transform, centering, standardizing), yielding a related, but new, variable. In order to be able to track the series of transfomations performed, these are recorded as the transformations vector. Each element is a structure with a name field (e.g., 'center') and a parameter field (e.g., one scalar for centering).

Navigate to:     General Description     Class Structure     Class Construction     Class Functions

Each field can be accessed by the dot (.) operation, or by the GET function. Most fields can be modified using the SET function, except for those designated as READ-ONLY which are computed automatically by the class methods and cannot be handled by the user. Some fields, which we found to be more commonly accessed, have specific SET/GET functions that are listed in the table.

		Field	Description	Type	Default	Dedicated Get/Set Function
		name	name of object, should be short and used as identifier. This field can never be empty.	string	'unnamed'
		description	verbal description of the class content.	string	''
		source	verbal description of the source of information.	string	''
	read only	no_samples	number of samples, n. Set to zero for void objects.	integer scalar	0	nosamples
		data	measured value of the variable for each of the n samples. Missing values are denoted by NaNs.	numerical vector	[]	() indexing (all values), or {} indexing (nonmissing values only)
		units	units of data.	string	''
		level	level of the variable.	unknown/nominal/ordinal/numerical	'unknown'
		lut	lookup table mapping integers to descriptions. Used for nominal variables only.	cell array of strings	{}	can be accessed using () indexing
	read only	transformations	keeps track of the transformation that were applied to the variable. Each transformation is a structure with fields 'name' and 'parameters'. Later, I plan to create a transformation object.	vector of transformation structures	[]
		minmax	allowed range of the variable. It is a structure with a field 'population' that holds the population range, and a field 'sample' that holds the sample range, and is read-only. If the population minimum and maximum are unknown, NaN is substituted.	a structure with two 2-vectors	[NaN NaN] for both population and sample minmax	min,max,minmax read sample values. pminmax reads population values.
		mean	mean of the variable. It is a structure with a field 'population' that holds the population mean, and a field 'sample' that holds the sample mean, and is read-only. If the population mean is unknown, NaN is substituted.	a structure with two scalars	NaN for both population and sample mean	mean for sample value, pmean for population value
		variance	variance of the variable. It is a structure with a field 'population' that holds the population variance, and a field 'sample' that holds the sample variance, and is read-only. If the population variance is unknown, NaN is substituted.	a structure with two scalars	NaN for both population and sample variance	var for sample value
		distribution	the distribution associated with the variable. This is a structure with fields 'name' and 'parameters'. Later, I plan to create a distribution object.	vector of distribution structures	name='unknown'; parameters=[]
	read only	no_missing	number of missing samples. Set to zero for void objects.	integer scalar	0	nomissing

Empty instance (scalar)

an empty variable instance, with all fields initialized to their default values.

syntax: vr = variable;

Empty instance (vector)

a vector of empty variable instances.

syntax: vr = variable(no_instances);

Copy constructor

a variable instance is copied into another.

syntax: vr_destination = variable(vr_origin);

Construction by field names

an instance is formed by directly providing field values. Any field which is not read-only is permitted.

syntax: vr = variable(field_name, field_value, ...);

example: vr = variable('name','demo variable', 'data',[1.1 1.2 1.3]);

Casting from a matrix

an instance is formed by providing a variable-by-sample matrix.

syntax: vr = variable(data);

Analysis:

detectoutliers

detects outliers

Characteristics of variable:

entropy

entropy of the variable

hist

computes and plots the histogram

Computations:

emutualentropy

estimates pairwise mutual entropy

Constructors:

loadobj

basic load function

variable

constructor method

Display functions:

display

display method

Indexing:

end

end keyword

instance

extracts a specific instance from a variable array

subsref

basic indexing method

Operators:

deletesamples

eliminate samples from a variable instance

== (eq)

element-wise logical operator

>= (ge)

element-wise logical operator

> (gt)

element-wise logical operator

<= (le)

element-wise logical operator

< (lt)

element-wise logical operator

- (minus)

subtracts two variables

/ (mrdivide)

divides a variable by a scalar

~= (ne)

element-wise logical operator

+ (plus)

adds two variables

./ (rdivide)

divides two variables

.* (times)

multiplies two variables

Queries:

iscomplete

checks for complete variable(s)

isnominal

checks for nominal variable(s)

isnumeric

checks for numeric variable(s)

isordinal

checks for ordinal variable(s)

isunknown

checks for unknown variable(s)

SET/GET functions:

get

get method

iqr

computes the inter-quartile range of variable(s)

max

retrieves the sample max of variable(s)

mean

retrieves the sample mean of variable(s)

median

retrieves the sample median of variable(s)

min

retrieves the sample min of variable(s)

minmax

retrieves the sample minmax of variable(s)

name

retrieves the name of variable(s)

nanless

extracts only complete data (without missing values)

nomissing

retrieves the number of missing values

nosamples

retrieves the number of samples in variable

pmean

retrieves the population mean of variable(s)

pminmax

retrieves the population minmax values of a variable

quantile

computes the quantiles of variable(s)

set

set method

std

retrieves the sample standard deviation of variable(s)

var

retrieves the sample variance of variable(s)

Transformations:

fillin

fills missing data

quantize

turns a numeric variable into a nominal one

transform

transforms variable data

Update functions:

computemean

calculates sample mean

computeminmax

calculates sample min and max values

computemissing

finds the number of missing values

computevariance

calculates sample variance