Files
df-research/dataframe/docs/StardustDocs/topics/operations/utils/any.md
2026-02-08 11:20:43 -10:00

71 lines
1.2 KiB
Markdown
Vendored

# any
<web-summary>
Discover `any` operation in Kotlin Dataframe.
</web-summary>
<card-summary>
Discover `any` operation in Kotlin Dataframe.
</card-summary>
<link-summary>
Discover `any` operation in Kotlin Dataframe.
</link-summary>
<!---IMPORT org.jetbrains.kotlinx.dataframe.samples.api.utils.AnySamples-->
Checks if there is at least one row in the [](DataFrame.md) that satisfies the predicate.
Returns `Boolean` — `true` if there is at least one row that satisfies the predicate, `false` otherwise.
```kotlin
df.any { rowCondition }
rowCondition: (DataRow) -> Boolean
```
**Related operations**: [](all.md), [](filter.md), [](single.md), [](count.md).
### Examples
<!---FUN notebook_test_any_3-->
```kotlin
df
```
<!---END-->
<inline-frame src="./resources/notebook_test_any_3.html" width="100%" height="500px"></inline-frame>
Check if any person `age` is greater than 21:
<!---FUN notebook_test_any_4-->
```kotlin
df.any { age > 21 }
```
<!---END-->
Output:
```text
false
```
Check if there is any person with `age` equal to 15 and `name` equal to "Alice":
<!---FUN notebook_test_any_5-->
```kotlin
df.any { age == 15 && name == "Alice" }
```
<!---END-->
Output:
```text
true
```