> For the complete documentation index, see [llms.txt](https://docs.myjuicebox.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.myjuicebox.io/building-reports/story-designer/slices/selections-as-variables.md).
# Selections as variables
Slice selections don't always have to filter down the page
By default, selections in a slice filter the data in downstream slices. You can change this default behavior by configuring slice selections.
You can configure the slice selections to act as:
1. **Filters**. Slice selections will filter downstream slices. (This is the default configuration.)
2. **Variables**. Slice selections will not filter downstream slices but will be available for use in defining ingredients.
3. **Filters and Variables**. Slices selections will filter downstream slices and be available for use in defining ingredients.
## Finding Available Variables
To see what variables are available to a slice, open the [debug information](/building-reports/story-designer/slices/debug.md), select the slice, and click on the **Variables** tab. This shows you:
* The variable names you can reference in formulas
* The current selected values
* The structure of the data (especially important for date ranges)
When a slice is configured to use selections as "Variables" or "Filters and Variables", you'll see selected values here. If the Variables tab shows `{}`, it means either:
* The slice isn't filtered to use selections as variables, or
* Nothing has been selected yet
## Using Variables
Variables are most useful when you need to reference selections within your ingredient definitions rather than relying on automatic filtering. A common use case is comparing data across two different time periods.
### Example: Comparing Sales Across Quarters
Let's say you want to compare sales between two quarters that users can select. You can't accomplish this with regular filtering because you need to reference both quarters simultaneously.
**Setup**:
1. Create two dimension ingredients, both with the field definition `quarter`:
1. Label the first "Baseline Quarter"
2. Label the second "Comparison Quarter"
2. Create slice(s) to display these dimensions. For example, two Tables slices (one for each dimension)
3. Configure each Table slice to require a selection ("Select one required") and to use selections as "Variables"
{% hint style="warning" %}
**Important:** For this comparison example, you must use "Variables" only, not "Filters and Variables". If you used "Filters and Variables", selecting a baseline quarter would filter the data and prevent you from seeing other quarters in the comparison selection. Variables-only mode keeps the selections independent.
{% endhint %}
4. Add a Chooser slice after the second Table slice and open the [debug information](/building-reports/story-designer/slices/debug.md) for the Chooser slice. Navigate to the Variables tab. You'll see something like:
```
{
...
"baseline_quarter": "2023 Q1",
"comparison_quarter": "2023 Q4",
...
}
```
The variable names (like `baseline_quarter` and `comparison_quarter`) are the ingredient IDs that Juicebox generates. You'll use the ingredient IDs when using variables in your measure formulas.
{% hint style="info" %}
**Note:** Your actual ingredient IDs will look different (e.g., `quarter_gqo74qoi`) but for this example we'll use `baseline_quarter` and `comparison_quarter` for clarity.
{% endhint %}
5. Create new Sales Delta and Sales % Change measures and add them to the Chooser slice.
#### Creating the Sales Delta Measure
Create a new measure ingredient with this field definition:
```
sum(if(quarter = "{{comparison_quarter}}", sales, 0))
- sum(if(quarter = "{{baseline_quarter}}", sales, 0))
```
This formula:
* Calculates sales for the comparison quarter
* Subtracts sales for the baseline quarter
* Returns the absolute change in sales
**Optional:** Add a dynamic label to make it clearer:
```
Change from {{baseline_quarter}} to {{comparison_quarter}}
```
#### Creating the Sales % Change Measure
Create another measure to calculate the percent change:
```
(sum(if(quarter = "{{comparison_quarter}}", sales, 0)) -
sum(if(quarter = "{{baseline_quarter}}", sales, 0))) /
sum(if(quarter = "{{baseline_quarter}}", sales, 0))
```
With a dynamic label:
```
% Change from {{baseline_quarter}} to {{comparison_quarter}}
```
{% hint style="success" %}
**Tip:** You don't need to wrap the division in `safe_divide()` - Juicebox automatically handles division by zero for you.
{% endhint %}
{% hint style="info" %}
**Note:** In this example, we configured the slices to require selections, so the formulas will always have values to work with. If you're using variables in a scenario where selections might be empty (optional filters), you'll need to handle empty states using the `default()` function - see the section on handling empty selection states below.
{% endhint %}
### Range Variables
Range selections (date ranges, number ranges) return an array with two elements: a start value and an end value. You can access these using array indexing:
**Basic syntax (when you know a range will always be selected):**
```
sum(if(joined_date BETWEEN
date("{{campaign_date__between_raw[0]['id']}}") AND
date("{{campaign_date__between_raw[1]['id']}}"),
revenue))
```
* `[0]` gets the start value
* `[1]` gets the end value
* `['id']` extracts the value
* `date()` wraps the value to ensure proper date type conversion (for date ranges)
### Handling Empty Selection States
If a selection might be empty, your formula will break when trying to reference an undefined variable. Use the `default()` function to provide fallback values:
**For string variables:**
```
sum(if(category = "{{category_filter|default('All')}}", sales, 0))
```
**For range variables (date ranges, number ranges):**
```
sum(if(joined_date BETWEEN
date("{{(campaign_date__between_raw|default([{'id':'1970-01-01'}], true)|first).id}}") AND
date("{{(campaign_date__between_raw|default([{'id':'2099-12-31'}], true)|last).id}}"),
revenue))
```
Breaking down the range syntax:
* `campaign_date__between_raw` - the base variable name
* `|default([{'id':'1970-01-01'}], true)` - provides a fallback value if nothing is selected
* `|first` - gets the start value (similar to `[0]`, but works within the filter chain)
* `|last` - gets the end value (similar to `[1]`, but works within the filter chain)
* `.id` - extracts the value
* `date()` - converts to proper date type (for date ranges)
{% hint style="info" %}
**Why use `|first` and `|last` instead of `[0]` and `[1]`?**\
When you use the `default()` function, you're using Jinja template syntax (the `|` pipe operator chains operations together). Within this syntax, you need to use `|first` and `|last` instead of array indexing. If you're NOT using `default()`, you can use the simpler `[0]` and `[1]` syntax.
{% endhint %}
### Limitations
* For multi-select slices, the variable contains only the **first** selected value. If multiple values are selected, only the first will be available in the variable.
* You can use variables in ingredient formulas, labels, and many other places throughout Juicebox. However, variable support isn't available everywhere yet. If you need variables in a location that's not currently supported, please [let us know](/getting-started/reach-out-to-us.md)!
---
# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.
## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.
Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter:
```
GET https://docs.myjuicebox.io/building-reports/story-designer/slices/selections-as-variables.md?ask=&goal=
```
`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.
Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
