In Sparkflows, we can use the ‘Windows Analytics’ processor to perform Windows Analytics on a dataset. It facilitates partitioning the dataset based on a selected column and applies windows functions such as first_value, last_value, lag and lead to compute value.
To use the ‘Windows Analytics’ Processor:
- Select a set of columns to partition the dataset in the Partition By field.
- Select a set of columns to Order dataset in the Order By field.
- Select Windows Function to be used to perform analysis.
- Select a column in Analytics Columns. Windows function selected would be applied to the selected column. It would be added as a new column to the output dataset.
- Enter a value for Window Offset. It would be used when a function is selected as ‘Lead’ or ‘Lag’. It would determine how many backward or forward rows data would be used.
For more information read the Sparkflows Documentation here: