Summary

The If Then Split block conditionally routes all data into one of two outputs — True or False — based on a single logical formula that is evaluated once over the entire dataset connected to the test input. The formula must return either true or false. The output branch that is not selected is halted (see below), so downstream blocks on that branch will not execute and there are no side effects.

How does it work?

• test (mandatory) — connect a dataset here. The formula is evaluated a single time on the complete dataset provided to test.
• data (optional) — if connected, the dataset coming from data (instead of test) is routed into the chosen output. test is still always used for evaluating the formula.
• Formula — configured in the block options. Must evaluate to a boolean (true or false) when applied to the whole test dataset (or a subset defined inside the formula).

Halting Behavior

The non-selected output branch is marked halted. Halted branches are not executed at all — any downstream blocks connected (directly or indirectly) to a halted branch will not run. Halted branches produce no side effects.

Options

Formula

Enter a single logical expression that is evaluated once on the complete test dataset. The expression must evaluate to true or false. Use Omniscope aggregate and SUBSET functions to compute dataset-level statistics or to evaluate conditions on specific subsets.

Examples Using Omniscope SUBSET Functions:

These formulas are dataset-level checks — they return a single boolean and therefore route all data to either True or False.

Example 1 — Route by average sales for a region

Route all data to True if the average sales in Region A exceed 100,000:

SUBSET_MEAN([Sales], SUBSET([Region], "Region A", "=")) > 100000

• If the expression is true, all data (from data if connected, otherwise from test) goes to True.
• If false, all data goes to False.

Example 2 — Route by unique customers in a category

Route to True if there are more than 1,000 unique customers in the "Premium" category:

SUBSET_UNIQUECOUNT([Customer ID], SUBSET([Category], "Premium", "=")) > 1000

Example 3 — Route by an intersection subset (region + channel)

Check whether the mean sales for online sales in Region A exceed 50,000:

SUBSET_MEAN( [Sales], INTERSECTION( SUBSET([Region], "Region A", "="), SUBSET([Channel], "Online", "=") ) ) > 50000

Notes on SUBSET usage

• SUBSET(field, value, op) defines a subset filter; these filters can be combined with INTERSECTION, UNION, etc.
• SUBSET_* aggregates (e.g., SUBSET_SUM, SUBSET_MEAN, SUBSET_MEDIAN, SUBSET_COUNT, SUBSET_UNIQUECOUNT) compute a single value for the chosen subset — ideal for dataset-level formulas used by this block.

Outputs

• True — receives all data when the formula evaluated on the test dataset is true. If the optional data input is connected, that dataset is routed here instead.
• False — receives all data when the formula evaluated on the test dataset is false. If data is connected, that dataset is routed here instead.
• The branch that is not chosen is halted — downstream blocks on that branch do not execute.

Examples summary

• The formula is evaluated once on the entire test dataset (or on a subset defined inside the formula).
• The boolean result routes all data to either True or False.
• Use SUBSET_* functions and subset combinators (INTERSECTION, UNION, etc.) to compute dataset-level metrics for the decision.