The aggregate node

The aggregate node lets you compute statistical functions on data_batches. This node can compute multiple statistic functions on multiple fields.

See nodes under /nodes/statistics for details (Not all of these are available here).

This node can only take data_batch items, so in most cases you will need some kind of batch or window node to collect batches.

The aggregate node produces a new stream of data consisting of data_point items.




.fields('val', 'val', 'val', 'val')
.functions('variance', 'sum', 'avg', 'count_distinct')
.as('variance', 'sum', 'average', 'count_distinct')



Parameter Description Default
fields( string_list ) names of the fields used for each computation
functions( string_list ) list of function names (see 'Functions' table below for possible values)
as( string_list ) names for the fields for output values defaults to the name of the compute function
keep( string_list ) fields that should be kept from the input data-items []


Name Description Note
sum computes the sum of all values numerical values only
avg computes the average of all values numerical values only
min computes the minimum value
max computes the maximum value
mean computes the mean value numerical values only
median computes the median value numerical values only
first outputs the first (oldest) value
last outputs the last (newest) value
count outputs the total number of values
count_distinct outputs the number of unique values
count_change outputs the number of value changes for the given field
range computes the range between minimum and maximum
variance computes the variance numerical values only
stddev computes the standard deviation of the values numerical values only
skew computes the skew of all values numerical values only