The input to calculations i.e. the data that the calculation will process can be provided either by features or by the output of other calculations in the same calculation algorithm. When an calculation algorithm is defined features or calculations are connected to the input pins of a calculation. The available input pins for a calculation are predefined by the calculation type.
There are two types of input pins. The type of the input pin defines if features or features and the output of other calculations can be connected to the input pin. Feature only input pins can not get their input from other calculations. Feature or calculation input pins can either be connected directly to features or to the output of another calculations.Feature only input pins cannot be connected to other calculated features.
Input pin constraints further define which features or calculations can be connected to an input pin of a calculation. This constraint is defined by the calculation type and therefore not modifiable. Table 1 lists the constraints that can be applied to input pins.
Depending on the constraint that is applied to an input pin the pin can allow multiple connections. I.e. more than one feature and/or calculations can be connected to that input pin. An input pin with a constraint that allows multiplicity will define a minimum and a maximum number of features and/or calculations that can be connected to that input pin. When multiplicity is allowed the sequence of connected features or calculations might be important for the calculation.
Constraint |
Notes |
Multiplicity possible |
None |
A feature of any type or a calculations with an output conforming to any type can be connected to that input pin |
No |
Feature type |
Only features of a specific feature type (predefined by the calculation type) or calculations that generate observations conforming to features of that specific feature type can be connected to the input pin. |
Yes |
Same feature types |
Any feature type can be connected to that input pin, all connected features have to be of the same feature type. The feature type is not predefined by the calculation but defined by the first feature or calculation connected to that pin. |
Yes |
Rules |
Custom rule that defines characteristics the feature's or calculations observations and its property structure must have (see Table 2) |
No |
Table1: Constraint types for input pins
If an input pin is constrained with the constraint type rules it will include one or more rules that define which characteristic the observations must have.
Rule |
Description |
|
Observation location type rule |
Only features that have observation location or calculations that conform to.. |
No |
Observation property structure rules |
Only features whose observation property structure conforms to the defined rules can be used at that input pin. Observation property structure rules, define which property types are needed at which level of the property structure. E.g. the rules defined for an input pin might require that the features or calculations connected to that input pin one have a numerical property on level one. When a feature or calculation output is connected to that pin the user will then have to specify which numerical property of level one shall be used by the calculation. |
No |
Table 2: Rules
