> For the complete documentation index, see [llms.txt](https://docs.agenticflow.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.agenticflow.ai/reference/nodes/delete_data.md).
# Delete Data
**Action ID:** `delete_dataset_rows`
## Description
Delete rows from a dataset that match filter conditions.
## Input Parameters
| Name | Type | Required | Default | Description |
| ---------- | -------- | :------: | ------- | --------------------------------------------------------------------- |
| dataset | dropdown | ✓ | - | The dataset to delete rows from. |
| conditions | array | - | \[] | Filter conditions to select rows to delete (empty = delete all rows). |
| logic | dropdown | - | and | How to combine filter conditions (AND/OR). Options: and, or |
### Filter Condition Structure
Each condition in the `conditions` array contains:
| Field | Type | Description |
| -------- | -------- | ----------------------------------------------------------------------------------------- |
| column | string | Column name to filter by |
| operator | dropdown | Comparison operator: equals, not\_equals, contains, greater\_than, less\_than, in |
| value | string | Value to compare against. For 'in' operator, provide comma-separated values or JSON array |
View JSON Schema
```json
{
"description": "Delete Dataset Rows node input.",
"properties": {
"dataset": {
"description": "The dataset to delete rows from.",
"title": "Dataset",
"type": "string"
},
"conditions": {
"default": [],
"description": "Filter conditions to select rows to delete (empty = delete all rows).",
"items": {
"properties": {
"column": {
"description": "Column name to filter by",
"title": "Column",
"type": "string"
},
"operator": {
"default": "equals",
"description": "Comparison operator",
"enum": ["equals", "not_equals", "contains", "greater_than", "less_than", "in"],
"title": "Operator",
"type": "string"
},
"value": {
"description": "Value to compare against. For 'in' operator, provide comma-separated values or JSON array",
"title": "Value",
"type": "string"
}
},
"required": ["column", "value"],
"type": "object"
},
"title": "Filter Conditions",
"type": "array"
},
"logic": {
"default": "and",
"description": "How to combine filter conditions (AND/OR).",
"enum": ["and", "or"],
"title": "Logic Operator",
"type": "string"
}
},
"required": ["dataset"],
"title": "DeleteDatasetRowsNodeInput",
"type": "object"
}
```
## Output Parameters
| Name | Type | Description |
| -------------- | ------- | --------------------------------- |
| deleted\_count | integer | Number of rows that were deleted. |
View JSON Schema
```json
{
"description": "Delete Dataset Rows node output.",
"properties": {
"deleted_count": {
"description": "Number of rows that were deleted.",
"title": "Deleted Count",
"type": "integer"
}
},
"required": ["deleted_count"],
"title": "DeleteDatasetRowsNodeOutput",
"type": "object"
}
```
## How It Works
This node deletes rows from a dataset based on filter conditions. It evaluates each row against the specified conditions using the chosen logic operator (AND/OR). Matching rows are permanently removed from the dataset. The node validates the dataset ID format (26-character ULID) before execution and returns the count of deleted rows. If no conditions are provided, all rows in the dataset will be deleted.
## Usage Examples
### Example 1: Delete Inactive Users
**Input:**
```
dataset: "01K8ZM9T72FNBZAGA629KZFXR5"
conditions: [
{
"column": "status",
"operator": "equals",
"value": "inactive"
}
]
logic: "and"
```
**Output:**
```
deleted_count: 47
```
### Example 2: Delete Old Records
**Input:**
```
dataset: "01K8ZM9T72FNBZAGA629KZFXR5"
conditions: [
{
"column": "created_date",
"operator": "less_than",
"value": "2023-01-01"
}
]
logic: "and"
```
**Output:**
```
deleted_count: 1250
```
### Example 3: Delete Multiple Status Types
**Input:**
```
dataset: "01K8ZM9T72FNBZAGA629KZFXR5"
conditions: [
{
"column": "status",
"operator": "in",
"value": "deleted,archived,expired"
}
]
logic: "and"
```
**Output:**
```
deleted_count: 328
```
### Example 4: Delete with Multiple Conditions (AND)
**Input:**
```
dataset: "01K8ZM9T72FNBZAGA629KZFXR5"
conditions: [
{
"column": "status",
"operator": "equals",
"value": "pending"
},
{
"column": "age_days",
"operator": "greater_than",
"value": "90"
}
]
logic: "and"
```
**Output:**
```
deleted_count: 156
```
### Example 5: Delete with OR Logic
**Input:**
```
dataset: "01K8ZM9T72FNBZAGA629KZFXR5"
conditions: [
{
"column": "email",
"operator": "contains",
"value": "spam"
},
{
"column": "valid",
"operator": "equals",
"value": "false"
}
]
logic: "or"
```
**Output:**
```
deleted_count: 89
```
## Common Use Cases
* **Data Cleanup**: Remove outdated, invalid, or obsolete records from datasets
* **User Management**: Delete inactive or suspended user accounts
* **Compliance**: Remove data past retention period for GDPR/privacy compliance
* **Quality Control**: Delete records that failed validation or processing
* **Batch Operations**: Remove multiple records matching specific criteria in one operation
* **Status-Based Deletion**: Clean up records in terminal states like "completed", "failed", or "cancelled"
* **Archive Management**: Remove archived records from active datasets
* **Test Data Cleanup**: Clear test or demo data from production datasets
## Error Handling
| Error Type | Cause | Solution |
| ------------------- | ------------------------------------ | ----------------------------------------------------------------- |
| Dataset Not Found | Dataset ID doesn't exist | Verify the dataset ID is correct and the dataset exists |
| Invalid Dataset ID | Dataset ID format is incorrect | Ensure dataset ID is a 26-character ULID |
| Dataset ID Required | Dataset parameter is empty | Provide a valid dataset ID |
| Invalid Column | Column name doesn't exist in dataset | Check the dataset schema for available column names |
| Invalid Operator | Operator not in allowed list | Use: equals, not\_equals, contains, greater\_than, less\_than, in |
| Malformed Condition | Missing required fields | Ensure each condition has column, operator, and value |
| Permission Denied | Insufficient permissions | Verify you have delete permissions on the dataset |
| Delete Failed | Server error during deletion | Retry the operation or check server logs |
## Notes
* **Permanent Operation**: Deleted rows cannot be recovered. Use with caution and consider backing up data first.
* **Empty Conditions Warning**: If conditions array is empty, ALL rows in the dataset will be deleted. Double-check before execution.
* **Validation**: The node validates dataset ID format (must be 26-character ULID) before attempting deletion.
* **Atomic Operation**: The delete operation is atomic, ensuring data consistency.
* **Performance**: Deleting large numbers of rows may take time. Monitor the deleted\_count in the output.
* **Filter First**: Consider using the Query Data node to preview matching rows before deletion.
* **Audit Trail**: Keep logs of delete operations for compliance and troubleshooting purposes.
* **Conditional Logic**: Use AND logic to narrow down deletion criteria, OR logic to broaden the scope.
* **Batch Processing**: For very large deletions, consider breaking them into smaller batches to avoid timeout issues.
* **Dependencies**: Ensure no other workflows or processes depend on the rows being deleted.
* **Cascade Effects**: Be aware of any relationships or references that might be affected by row deletion.