This is an old revision of the document!
Table of Contents
FILTER BY CONDITION
Category: Transform / Filters
Description
This action calculates an expression for every row (without adding a new column) and keeps only rows where this expression is TRUE. All other rows are removed.
The expression must be a valid condition – it must return either TRUE or FALSE for any row in the table. If it results in a non-boolean value (i.e. not TRUE or FALSE) for at least one row it will cause a project execution error.
Use cases
Create subsets of a dataset based on values across multiple columns.
Clean a dataset by removing records that don't meet certain criteria, or evaluate to a certain value.
Action settings
Setting | Description |
---|---|
Expression | Enter an expression that evaluates to a boolean TRUE/FALSE result. The expression will be applied to each record and remove records that do not meet the criteria (evaluate to FALSE). |
Remarks
Examples of valid expressions:
[Amount]=0 [Length] * [Width] * [Height] <= 1000 contains([City], 'York') not isempty([Email])
If no rows satisfy the condition then all rows are removed.
Examples
Objective: Find all rivers that are longer than 6000 km.
Source table: The longest rivers in the world
River | Length (km) | Continent |
---|---|---|
Nile | 6650 | Africa |
Amazon | 6400 | South America |
Yangtze | 6300 | Asia |
Mississippi | 6275 | North America |
Yenisei | 5539 | Asia |
Yellow River | 5464 | Asia |
Ob | 5410 | Asia |
Paraná | 4880 | South America |
Action parameters:
Expression: [Length (km)] > 6000
Result:
River | Length (km) | Continent |
---|---|---|
Nile | 6650 | Africa |
Amazon | 6400 | South America |
Yangtze | 6300 | Asia |
Mississippi | 6275 | North America |