This window is a moving (sliding) length window extending the specified number of elements into the past. The window takes a single expression as a parameter providing a numeric size value that defines the window size:

length(size_expression)

The below example sums the price for the last 5 stock ticks for symbol GE.

select sum(price) from StockTickEvent(symbol='GE')#length(5)

The next example keeps a length window of 10 events of stock trade events, with a separate window for each symbol. The sum of price is calculated only for the last 10 events for each symbol and aggregates per symbol:

select sum(price) from StockTickEvent#groupwin(symbol)#length(10) group by symbol

A length window of 1 is equivalent to the last event window lastevent. The lastevent data window is the preferred notation:

select * from StockTickEvent#lastevent	// Prefer this
// ... equivalent to ...
select * from StockTickEvent#length(1)

This window buffers events (tumbling window) and releases them when a given minimum number of events has been collected. Provide an expression defining the number of events to batch as a parameter:

length_batch(size_expression)

The next statement buffers events until a minimum of 10 events have collected. Listeners to updates posted by this window receive updated information only when 10 or more events have collected.

select * from StockTickEvent#length_batch(10)

Similar to the time window, this window is a moving (sliding) time window extending the specified time interval into the past, but based on the long-type time value supplied by a timestamp expression. The window takes two parameters: the expression to return long-typed timestamp values, and a time period or expression that provides a number of seconds:

ext_timed(timestamp_expression, time_period)

ext_timed(timestamp_expression, seconds_interval_expression)

The key difference comparing the externally-timed window to the regular time window is that the window slides not based on the runtime time, but strictly based on the result of the timestamp expression when evaluated against the events entering the window.

The algorithm underlying the window compares the timestamp value returned by the expression when the oldest event arrived with the timestamp value returned by the expression for the newest arriving event on event arrival. If the time interval between the timestamp values is larger then the timer period parameter, then the algorithm removes all oldest events tail-first until the difference between the oldest and newest event is within the time interval. The window therefore slides only when events arrive and only considers each event's timestamp property (or other expression value returned) and not runtime time.

This window holds stock tick events of the last 10 seconds based on the timestamp property in StockTickEvent.

select * from StockTickEvent#ext_timed(timestamp, 10 seconds)

The externally-timed data window expects strict ordering of the timestamp values returned by the timestamp expression. The window is not useful for ordering events in time order, please use the time-order window instead.

This window buffers events (tumbling window) and releases them every specified time interval in one update. The window takes a time period or an expression providing a number of seconds as a parameter, plus optional parameters described next.

time_batch(time_period [,optional_reference_point] [,flow_control])

time_batch(seconds_interval_expression [,optional_reference_point] [,flow_control])

The time batch window takes a second, optional parameter that serves as a reference point to batch flush times. If not specified, the arrival of the first event into the batch window sets the reference point. Therefore if the reference point is not specified and the first event arrives at time t₁, then the batch flushes at time t₁ plus time_period and every time_period thereafter.

Similar to the time batch window, this window buffers events (tumbling) and releases them every specified time interval in one update, but based on the long-type time value supplied by a timestamp expression. The window has two required parameters taking an expression that returns long-typed timestamp values and a time period or constant-value expression that provides a number of seconds:

ext_timed_batch(timestamp_expression, time_period [,optional_reference_point])

ext_timed_batch(timestamp_expression, seconds_interval_expression [,optional_reference_point])

The externally-timed batch window takes a third, optional parameter that serves as a reference point to batch flush times. If not specified, the arrival of the first event into the batch window sets the reference point. Therefore if the reference point is not specified and the first event arrives at time t₁, then the batch flushes at time t₁ plus time_period and every time_period thereafter.

The key difference comparing the externally-timed batch window to the regular time batch window is that the window tumbles not based on the runtime time, but strictly based on the result of the timestamp expression when evaluated against the events entering the window.

The algorithm underlying the window compares the timestamp value returned by the expression when the oldest event arrived with the timestamp value returned by the expression for the newest arriving event on event arrival. If the time interval between the timestamp values is larger then the timer period parameter, then the algorithm posts the current batch of events. The window therefore posts batches only when events arrive and only considers each event's timestamp property (or other expression value returned) and not runtime time.

Note that using this window means that the runtime keeps events in memory until the time is up: Consider your event arrival rate and determine if this is the behavior you want. Use context declaration or output rate limiting such as output snapshot as an alternative.

The below example batches events into a 5 second window releasing new batches every 5 seconds. Listeners to updates posted by this window receive updated information only when event arrive with timestamps that indicate the start of a new batch:

select * from StockTickEvent#ext_timed_batch(timestamp, 5 sec)

The optional reference point is provided as a long-value of milliseconds (or microseconds) relative to January 1, 1970 and time 00:00:00.

The following example statement sets the reference point to 5 seconds and the batch size to 1 hour, so that each batch output is 5 seconds after each hour:

select * from OrderSummaryEvent#ext_timed_batch(timestamp, 1 hour, 5000L)

The externally-timed data window expects strict ordering of the timestamp values returned by the timestamp expression. The window is not useful for ordering events in time order, please use the time-order window instead.

This data window is a combination of time and length batch (tumbling) windows. Similar to the time and length batch windows, this batches events and releases the batched events when either one of the following conditions occurs, whichever occurs first: the data window has collected a given number of events, or a given time interval has passed.

The parameters take 2 forms. The first form accepts a time period or an expression providing a number of seconds, and an expression for the number of events:

time_length_batch(time_period, number_of_events_expression)

time_length_batch(seconds_interval_expression, number_of_events_expression)

The next example shows a time-length combination batch window that batches up to 100 events or all events arriving within a 1-second time interval, whichever condition occurs first:

 select * from MyEvent#time_length_batch(1 sec, 100)

In this example, if 100 events arrive into the window before a 1-second time interval passes, the window posts the batch of 100 events. If less then 100 events arrive within a 1-second interval, the window posts all events that arrived within the 1-second interval at the end of the interval.

By default, if there are no events arriving in the current interval (insert stream), and no events remain from the prior batch (remove stream), then the window does not post results to listeners. This window allows overriding this default behavior via flow control keywords.

The synopsis of the window with flow control parameters is:

time_length_batch(time_period or seconds_interval_expression, number_of_events_expression, 
    "flow control keyword [, keyword...]")

The FORCE_UPDATE flow control keyword instructs the window to post an empty result set to listeners if there is no data to post for an interval. The window begins posting no later then after one time interval passed after the first event arrives. When using this keyword the irstream keyword should be used in the select clause to ensure the remove stream is also output.

The START_EAGER flow control keyword instructs the window to post empty result sets even before the first event arrives, starting a time interval at statement deployment time. As when using FORCE_UPDATE, the window also posts an empty result set to listeners if there is no data to post for an interval, however it starts doing so at time of statement deployment rather then at the time of arrival of the first event.

Taking the two flow control keywords in one sample statement, this example presents a window that waits for 10 seconds or reacts when the 5th event arrives, whichever comes first. It posts empty result sets after one interval after the statement gets deployed and keeps posting an empty result set as no events arrive during intervals:

 select * from MyEvent#time_length_batch(10 sec, 5, "FORCE_UPDATE, START_EAGER")

This data window is a specialized moving (sliding) time window that differs from the regular time window in that it accumulates events until no more events arrive within a given time interval, and only then releases the accumulated events as a remove stream.

The window accepts a single parameter: the time period or seconds-expression specifying the length of the time interval during which no events must arrive until the window releases accumulated events. The synopsis is as follows:

time_accum(time_period)

time_accum(seconds_interval_expression)

The next example shows a time-accumulating window that accumulates events, and then releases events if within the time interval no more events arrive:

 select * from MyEvent#time_accum(10 sec)

This example accumulates events, until when for a period of 10 seconds no more MyEvent events arrive, at which time it posts all accumulated MyEvent events.

Your application may only be interested in the batches of events as events leave the data window. This can be done simply by selecting the remove stream of this data window, populated by the runtime as accumulated events leave the data window all-at-once when no events arrive during the time interval following the time the last event arrived:

 select rstream * from MyEvent#time_accum(10 sec)

If there are no events arriving, then the window does not post results to listeners.

This keep-all data window simply retains all events. The window does not remove events from the data window, unless used with a named window and the on delete clause.

The window accepts no parameters. The synopsis is as follows:

keepall

The next example shows a keep-all window that accumulates all events received into the window:

 select * from MyEvent#keepall

Note that since the window does not release events, care must be taken to prevent retained events from using all available resources.

The firstlength window retains the very first size_expression events.

The synopsis is:

firstlength(size_expression)

If used within a named window and an on-delete clause deletes events, the window accepts further arriving events until the number of retained events reaches the size of size_expression.

The below example creates a window that retains only the first 10 events:

select * from MyEvent#firstlength(10)

The firsttime window retains all events arriving within a given time interval after statement start.

The synopsis is:

firsttime(time_period)

firsttime(seconds_interval_expression)

The below example creates a window that retains only those events arriving within 1 minute and 10 seconds of statement start:

select * from MyEvent#firsttime(1 minute 10 seconds)

The expr data window applies an expiry expression and removes events from the data window when the expression returns false.

Use this window to implement rolling and dynamically shrinking or expanding time, length or other windows. Rolling can, for example, be controlled based on event properties of arriving events, based on aggregation values or based on the return result of user-defined functions. Use this window to accumulate events until a value changes or other condition occurs based on arriving events or change of a variable value.

The synopsis is:

expr(expiry_expression)

The expiry expression can be any expression including expressions on event properties, variables, aggregation functions or user-defined functions. The window applies this expression to the oldest event(s) currently in the window, as described next.

When a new event arrives or when a variable value referenced by the expiry expression changes then the window applies the expiry expression starting from the oldest event in the data window. If the expiry expression returns false for the oldest event, the window removes the event from the data window. The window then applies the expression to the next oldest event. If the expiry expression returns true for the oldest event, no further evaluation takes place and the window indicates any new and expired events through insert and remove stream.

By using variables in the expiry expression it is possible to change the behavior of the window dynamically at runtime. When one or more variables used in the expression are updated the window evaluates the expiry expression starting from the oldest event.

Aggregation functions, if present in the expiry expression, are continuously updated as events enter and leave the data window. Use the grouped data window with this window to compute aggregations per group.

The runtime makes the following built-in properties available to the expiry expression:

The expr_batch buffers events (tumbling window) and releases them when a given expiry expression returns true.

Use this window to implement dynamic or custom batching behavior, such as for dynamically shrinking or growing time, length or other batches, for batching based on event properties of arriving events, aggregation values or for batching based on a user-defined function.

The synopsis is:

expr_batch(expiry_expression, [include_triggering_event])

The expiry expression can be any expression including expressions on event properties, variables, aggregation functions or user-defined functions. The window applies this expression to arriving event(s), as described next.

The optional second parameter include_triggering_event defines whether to include the event that triggers the batch in the current batch (true, the default) or in the next batch (false).

When a new event arrives or when a variable value referenced by the expiry expression changes or when events get removed from the data window then the window applies the expiry expression. If the expiry expression returns true the data window posts the collected events as the insert stream and the last batch of events as remove stream.

By using variables in the expiry expression it is possible to change the behavior of the window dynamically at runtime. When one or more variables used in the expression are updated the window evaluates the expiry expression as well.

Aggregation functions, if present in the expiry expression, are continuously updated as events enter the data window and reset when the runtime posts a batch of events. Use the grouped data window with this window to compute aggregations per group.

The compiler makes the following built-in properties available to the expiry expression:

The unique window is a window that includes only the most recent among events having the same value(s) for the result of the specified expression or list of expressions.

The synopsis is:

unique(unique_expression [, unique_expression ...])

The window acts as a length window of size 1 for each distinct value returned by an expression, or combination of values returned by multiple expressions. It thus posts as old events the prior event of the same value(s), if any.

Specifying #groupwin groups events into sub-data-window by the value returned by the specified expression or the combination of values returned by a list of expressions. The #groupwin takes a single expression to supply the group criteria values, or a list of expressions as parameters, as the synopsis shows:

groupwin(grouping_expression [, grouping_expression ...])

This window exposes the last element:

lastevent

The window acts as a length window of size 1. It thus posts as old events the prior event in the stream, if any.

This example statement retains the last stock tick event for the symbol GE.

select * from StockTickEvent(symbol='GE')#lastevent

This window retains only the first arriving event:

firstevent

All events arriving after the first event are discarded.

If used within a named window and an on-delete clause deletes the first event, the window resets and will retain the next arriving event.

An example of a statement that retains the first ReferenceData event arriving is:

select * from ReferenceData#firstevent

The firstunique window retains only the very first among events having the same value for the specified expression or list of expressions.

The synopsis is:

firstunique(unique_expression [, unique_expression ...])

This window sorts by values returned by the specified expression or list of expressions and keeps only the top (or bottom) events up to the given size.

This window retains all events in the stream that fall into the sort range. Use the ranked window as described next to retain events per unique key(s) and sorted.

The syntax is as follows:

sort(size_expression, 
    sort_criteria_expression [asc/desc][, sort_criteria_expression [asc/desc]...]) 

An expression may be followed by the optional asc or desc keywords to indicate that the values returned by that expression are sorted in ascending or descending sort order.

The window below retains only those events that have the highest 10 prices considering all events (and not only the last event per symbol, see rank below) and reports a total price:

select sum(price) from StockTickEvent#sort(10, price desc)

The following example sorts events first by price in descending order, and then by symbol name in ascending (alphabetical) order, keeping only the 10 events with the highest price (with ties resolved by alphabetical order of symbol).

select * from StockTickEvent#sort(10, price desc, symbol asc)

The sorted window is often used with the prev, prevwindow or prevtail single-row functions to output properties of events at a certain position or to output the complete data window according to sort order.

Use the grouped window to retain a separate sort window for each group. For example, the windows groupwin(market)#sort(10, price desc) instruct the runtime to retain, per market, the highest 10 prices.

This window retains only the most recent among events having the same value for the criteria expression(s), sorted by sort criteria expressions and keeps only the top events up to the given size.

This window is similar to the sorted window in that it keeps only the top (or bottom) events up to the given size, however the window also retains only the most recent among events having the same value(s) for the specified uniqueness expression(s).

The syntax is as follows:

rank(unique_expression [, unique_expression ...],
    size_expression, 
    sort_criteria_expression [asc/desc][, sort_criteria_expression [asc/desc]...]) 

This window orders events that arrive out-of-order, using timestamp-values provided by an expression, and by comparing that timestamp value to runtime time.

The syntax for this window is as follows.

time_order(timestamp_expression, time_period)

time_order(timestamp_expression, seconds_interval_expression)

The first parameter to the window is the expression that supplies timestamp values. The timestamp is expected to be a long-typed value that denotes an event's time of consideration by the window (or other expression). This is typically the time of arrival. The second parameter is a number-of-seconds expression or the time period specifying the time interval that an arriving event should maximally be held, in order to consider older events arriving at a later time.

Since the window compares timestamp values to runtime time, the window requires that the timestamp values and runtime time are both following the same clock. Therefore, to the extend that the clocks that originated both timestamps differ, the window may produce inaccurate results.

As an example, the next statement uses the arrival_time property of MyTimestampedEvent events to order and release events by arrival time:

insert rstream into ArrivalTimeOrderedStream
select rstream * from MyTimestampedEvent#time_order(arrival_time, 10 sec)

In the example above, the arrival_time property holds a long-typed timestamp value. On arrival of an event, the runtime compares the timestamp value of each event to the tail-time of the window. The tail-time of the window is, in this example, 10 seconds before runtime time (continuously sliding). If the timestamp value indicates that the event is older then the tail-time of the time window, the event is released immediately in the remove stream. If the timestamp value indicates that the event is newer then the tail-time of the window, the window retains the event until runtime time moves such that the event timestamp is older then tail-time.

The examples thus holds each arriving event in memory anywhere from zero seconds to 10 seconds, to allow for older events (considering arrival time timestamp) to arrive. In other words, the window holds an event with an arrival time equal to runtime time for 10 seconds. The window holds an event with an arrival time that is 2 seconds older then runtime time for 8 seconds. The window holds an event with an arrival time that is 10 or more seconds older then runtime time for zero seconds, and releases such (old) events immediately into the remove stream.

The insert stream of this sliding window consists of all arriving events. The remove stream of the window is ordered by timestamp value: The event that has the oldest timestamp value is released first, followed by the next newer events. Note the statement above uses the rstream keyword in both the insert into clause and the select clause to select ordered events only. It uses the insert into clause to makes such ordered stream available for subsequent statements to use.

It is up to your application to populate the timestamp property into your events or use a sensible expression that returns timestamp values for consideration by the window. The window also works well if you use externally-provided time via timer events.

This window retains events until runtime time reaches the value returned by the given timestamp expression.

The syntax for this window is as follows:

timetolive(timestamp_expression)

The only parameter to the window is the expression that supplies timestamp values. The timestamp is expected to be a long-typed value that denotes an event's time-to-live.

Since the window compares timestamp values to runtime time, the window requires that the timestamp values and runtime time are both following the same clock.

On arrival of an event, the runtime evaluates the timestamp expression and obtains a long-type timestamp. The runtime compares that timestamp value to runtime time:

As an example, the next statement uses the arrival_time property of MyTimestampedEvent events to release events by arrival time:

insert rstream into ArrivalTimeOrderedStream
select rstream * from MyTimestampedEvent#timetolive(arrival_time)

For example, assume runtime time is 8:00:00 (8 am).

The runtime evaluates the expression only once at the arrival of each event to determine that event's time-to-live. The time-to-live data structure organizes events by the value returned by the expression on event arrival into the data window.

The insert stream of this sliding window consists of all arriving events. The remove stream of the window is ordered by timestamp value: The event that has the oldest timestamp value is released first, followed by the next newer events. Note the statement above uses the rstream keyword in both the insert into clause and the select clause to select ordered events only. It uses the insert into clause to makes such ordered stream available for subsequent statements to use.

It is up to your application to populate the timestamp property into your events or use a sensible expression that returns timestamp values for consideration by the window. The window also works well if you use externally-provided time via timer events and if you have runtime time track watermarks.

The time-to-live data window is fully equivalent to the time-order data window with a zero value for the time period.

This window posts the number of events received from a stream or window plus any additional event properties or expression values listed as parameters. The synopsis is:

size([expression, ...] [ * ])

The window posts a single long-typed property named size. The window posts the prior size as old data, and the current size as new data to update listeners of the window. Via the iterator method of the statement the size value can also be polled (read). The window only posts output events when the size count changes and does not stay the same.

As optional parameters the window takes a list of expressions that the window evaluates against the last arriving event and provides along the size field. You may also provide the * wildcard selector to have the window output all event properties.

An alternative to receiving a data window event count is the prevcount function. Compared to the size window the prevcount function requires a data window while the size window does not. The related count(...) aggregation function provides a count per group when used with group by.

When combined with a data window, the size window reports the current number of events in the data window in the insert stream and the prior number of events in the data window as the remove stream. This example reports the number of tick events within the last 1 minute:

select size from StockTickEvent#time(1 min)#size

To select additional event properties you may add each event property to output as a parameter to the window.

The next example selects the symbol and feed event properties in addition to the size property:

select size, symbol, feed from StockTickEvent#time(1 min)#size(symbol, feed)

This example selects all event properties in addition to the size property:

select * from StockTickEvent#time(1 min)#size(*)

The size window is also useful in conjunction with a groupwin grouped-window to count the number of events per group. The EPL below returns the number of events per symbol.

select size from StockTickEvent#groupwin(symbol)#size

When used without a data window, the window simply counts the number of events:

select size from StockTickEvent#size

All windows can be used with pattern statements as well. The next EPL snippet shows a pattern that looks for tick events followed by trade events for the same symbol. The size window counts the number of occurrences of the pattern.

select size from pattern[every s=StockTickEvent -> TradeEvent(symbol=s.symbol)]#size

This window calculates univariate statistics on a numeric expression. The window takes a single value expression as a parameter plus any number of optional additional expressions to return properties of the last event. The value expression must return a numeric value:

uni(value_expression [,expression, ...] [ * ])

After the value expression you may optionally list additional expressions or event properties to evaluate for the stream and return their value based on the last arriving event. You may also provide the * wildcard selector to have the window output all event properties.

This window calculates regression and related intermediate results on the values returned by two expressions. The window takes two value expressions as parameters plus any number of optional additional expressions to return properties of the last event. The value expressions must return a numeric value:

linest(value_expression, value_expression [,expression, ...] [ * ])

After the two value expressions you may optionally list additional expressions or event properties to evaluate for the stream and return their value based on the last arriving event. You may also provide the * wildcard selector to have the window output all event properties.

This window calculates the correlation value on the value returned by two expressions. The window takes two value expressions as parameters plus any number of optional additional expressions to return properties of the last event. The value expressions must be return a numeric value:

correl(value_expression, value_expression [,expression, ...] [ * ])

After the two value expressions you may optionally list additional expressions or event properties to evaluate for the stream and return their value based on the last arriving event. You may also provide the * wildcard selector to have the window output all event properties.

This window returns the weighted average given an expression returning values to compute the average for and an expression returning weight. The window takes two value expressions as parameters plus any number of optional additional expressions to return properties of the last event. The value expressions must return numeric values:

weighted_avg(value_expression_field, value_expression_weight [,expression, ...] [ * ])

After the value expression you may optionally list additional expressions or event properties to evaluate for the stream and return their value based on the last arriving event. You may also provide the * wildcard selector to have the window output all event properties.