Expression filter

From Piki

The expression filter adds new features to a data unit based on an algebraic expression.

Expression
Name	Expression
Deployable	Yes
Static	Yes/No
Sample modifier	No
Feature modifier	Yes

Contents 1 Usage 1.1 Operators 1.2 Referencing data 1.3 Examples 2 Settings 3 See also

Usage

The expression filter creates one or more new features from a set of algebraic expressions. Each expression is evaluated for each sample in the data unit. The expression filter can, among other things, be used to delay a feature, split a feature up or create compound features from other features.

An expression consists of numbers referenced features and operators.

Operators

The available operators are the following:

+, -, *, /, %, ^, <, >, >=, <=, !=, ==, &, |, <<, >>, 'int', 'sqrt'

Boolean expressions evaluate to 1 for True and 0 for False. The special numbers Pi, e and eps (epsilon) can also be used.

Referencing data

The source feature can be accessed through the letter ‘d’. Other features are accessed in one of three ways:

• By name within brackets.
• By relative position within curled brackets.
• By absolute position (counted from one) preceded by @.

Internally the two last methods are converted to the first one automatically.

To reference other samples than the current use a ':' followed by + or - an offset. Not using + or - is treated as an absolute sample number.

'int' is used to cast to integers. (2.7)int -> 2 '>>' is used as much greater than, a >> b is true if a is 100 times greater than b

Examples

Task	Expression
Divide the selected column by 2	d/2
Delay the selected column samples one step (t-1)	{0:-1}
Advance the second column samples three steps (t+3)	@2:+3
Divide the column Price and the column Area	[Price]/[Area]
Set all values > 100 to 10 in the selected column.	10*(d>100)
Add the column Price1 and Price2 when Price3 equals 100.	([Price1] + [Price2]) * ([Price3] == 100])
Extract the second lowest digit of an integer number.	(d/10)int%10
Calculate a hypotenuse using Pythagoras' theorem	( ([A)^2) + ([B]^2) )sqrt

Settings

The settings can be modified using the settings browser.

Expression filter settings

(Source) Name: The source feature. The current sample of this feature can be accessed by the letter 'd'. Missing values: The value that missing values are replaced with. (New Expression) Name: The name of the new features created by the expression. Expression: The expression that is to be evaluated to create the new feature. Add: Add the expression to the list of expressions. Help: Displays a window with information on syntax and usage. (Remove Expression) Feature: A drop down list of expressions used for removal. Select the feature you want to remove and click Remove. Remove: Remove the feature selected in the Feature drop down. (Expression <*>)- A new category is created for each new expression added. Column: The name of the feature created by the expression. Expression: The expression used to create the new feature.

See also

• Filter - Article covering general filter principles.
• List of Filter components - List of all available filters.
• Data unit - Article explaining data units in detail.
• Data - Article explaining essential data concepts.