Skip to main content

计算数据

Synopsis

This operator calculates summaries for one or more PI data points via the PI Web API. Alternatively, a single performance equation can be evaluated.

Input

connection

This input port expects a Connection object if any. See the parameter connection entry for more information.

Output

output

This output port delivers calculated time series for the specified data points or performance equation.

connection

This output port delivers the Connection object from the input port. If the input port is not connected the port delivers nothing.

Parameters

Connection entry

This parameter is used to specify a repository location that represents a connection entry. Theconnectioncan also be provided using the connection input port.

Root path

Overrides the root path (data server) specified in the connection.

Query type

Specifies whether to look up a list of data items or to evaluate a performance equation.

Data item

The tag of the data item to look up.

Additional data item

An optional list of additional tags of data items to look up.

Client side batching

Enabling client-side batching splits the request into a sequence of smaller requests. Client-side batching allows to pull data sets larger than the PI Web API limit for single requests (often set to 150000 values), but adds overhead for requests below this limit.

Expression

The performance equation to be evaluated.

Start time

The start time of the period to calculate data for. Supports both date times and relative intervals. The default behavior is to start one day ago ('*-1d').

End time

The end time of the period to calculate data for. Supports both date times and relative intervals. The default behavior is to calculate data until the moment the operator is executed ('*').

Time zone

Specifies the time zone to use for evaluating the given start and end time.

Filter expression

An optional filter expression. Expression variables are relative to the data point. Use '.' to reference the containing attribute.

Time interval

The sampling interval. The default is '1h'.

Calculation mode

The type of calculation used for the summaries. For non-numeric data points, only 'count' is supported.

  • total: A totalization over the time range.
  • average: The average value over the time range.
  • minimum: The minimum value over the time range.
  • maximum: The maximum value over the time range.
  • range: The range value over the time range (minimum-maximum).
  • standard deviation: The standard deviation over the time range.
  • population standard deviation: The population standard deviation over the time range.
  • count: The sum of event count over the time range when calculation basis is event weighted. The sum of event time duration over the time range when calculation basis is time weighted.
  • percent good: Percent of data with good value during the calculation period. For time weighted calculations, the percentage is based on time. For event weighted calculations, the percent is based on event count.
  • percent good: Percent of data with good value during the calculation period. For time weighted calculations, the percentage is based on time. For event weighted calculations, the percent is based on event count.

Calculation mode

Specifies the method of evaluating the data over the time range.

  • time weighted: Weight the values in the calculation by the time over which they apply. Interpolation is based on whether the attribute is stepped. Interpolated events are generated at the boundaries if necessary.
  • time weighted (continuous): Apply weighting as in 'time weighted', but do all interpolation between values as if they represent continuous data, (standard interpolation) regardless of whether the attribute is stepped.
  • time weighted (discrete):应用权重加权,但插值函数olation between values is performed as if they represent discrete, unrelated values (stair step plot) regardless of the attribute is stepped.
  • event weighted: Evaluate values with equal weighting for each event. There must be at least one event within the time range to perform a successful calculation. Two events are required for standard deviation. In handling events at the boundary of the calculation, the following rules are applied: Use events at both boundaries when there is only one calculation interval. Include events at start time in multiple intervals and the intervals are in ascending time order. Include events at the end time in multiple intervals and the intervals are in descending time order.
  • event weighted (exclude most recent): The calculation behaves the same as 'event weighted', except in the handling of events at the boundary of summary intervals in a multiple intervals calculation. Use this option to prevent events at the intervals boundary from being double count at both intervals. With this option, events at the end time (most recent time) of an interval is not used in that interval.
  • event weighted (exclude earliest): Similar to the option 'event weighted (exclude most recent)'. Events at the start time (earliest time) of an interval is not used in that interval.
  • event weighted (include both ends): Events at both ends of the interval boundaries are included in the event weighted calculation.