# Waterfall
The Waterfall step allows to generate a data structure useful to build waterfall charts. It breaks down the variation between two values (usually between two dates) accross entities. Entities are found in the `labelsColumn`, and can optionally be regrouped under common parents found in the `parentsColumn` for drill-down purposes.
### Step parameters
* `Value column name`: the name of the value column tht will be used to compute waterfall blocks values
* `Column incl. start and end labels (usually dates)`: name of the column that includes the labels of the starting and ending blocks of the waterfall (usually dates)
* `Starting block label`: the label of the starting block to find in the column above
* `Ending block label`: the label of the ending block to find in the column above
* `Labels columns (for intermediate blocks)`: the name of the column where intermediate blocks labels are to be found (used to break down the total variation between the starting and ending block)
* `(Optional) Parents labels column (for drill-down)`: the name of the column where parent labels are to be found if any, for drill-down purposes (parents of the labels specified above)
* `(Optional) Group waterfall by`: the columns used to group waterfall computation (useful to keep track of a column used for filtering for example)
* `Sort by`: whether to sort waterfall blocks depending on their labels or on their values
* `Sort order`: whether to use ascending or descending order when sorting
### Example
**Input**
