How does TrendMiner determine tag type?
When using a supported TrendMiner connector, the tag types and interpolations used in the historian will be converted to one of the tag types supported by TrendMiner.
For ODBC data sources, the tag type mapping can be configured as explained in the linked articles.
For other connectivity options to common data historians TrendMiner uses the type mappings as listed in the tables below:
Aveva Pi
Historian Type
Mapped TrendMiner Tag Type
OLEDB D, SDI PIPointType.Digital
DIGITAL
OLEDB S, SDI PIPointType.String
STRING
OLEDB others
ANALOG
PIPoint.Step Property = True
DISCRETE
IP21
A special query is executed when using the plug-&-play connector that gets information if the interpolation type is stepped and information regarding the value’s ‘display length’. Given this, TrendMiner will assign any one of the supported tag types. The default mapping for different combinations of settings in InfoPlus to the options available in TrendMiner is as follows:
Table 15.IP_STEPPED
IP_TREND_VALUE[1]
IP_VALUE_FORMAT -> #_OF_SELECTIONS
IP_VALUE_FORMAT -> DISPLAY_LENGTH
TrendMiner tag type
>0
>0
DISCRETE
>0
DIGITAL
NOT NULL
AND NOT (NULL_As_Numeric = True && Empty String)
AND NOT (REAL or INTEGER)
STRING
Stepped
DISCRETE
All other cases
ANALOG
Aveva Wonderware
Historian Type
Mapped TrendMiner Tag Type
DoubleByteString, SingleByteString, Event, Structure
STRING
Double, Float
ANALOG
Considering RawType (Int1, Int2, Int4, UInt2, UInt4) and TagInterpolationType property
DISCRETE or ANALOG
GE Proficy
Historian Type
Mapped TrendMiner Tag Type
FixedString, VariableString, Time
STRING
Bool
DISCRETE
Considering StepValue (true or false) property
ANALOG (false); DISCRETE (true)
Exaquantum
Historian Type
Mapped TrendMiner Tag Type
String
STRING
Integer/long and discrete aggregation
DISCRETE
Others
ANALOG
PHD
Historian Type
Mapped TrendMiner Tag Type
U (unicode)
STRING
C (char)
STRING
I (integer)
DISCRETE
L (long)
DISCRETE
Others
ANALOG