Using the Python SDK
The Athina Python SDK provides a simple interface to update cells in your datasets:
from athina_client.dataset import Dataset
# Define the cells you want to update
cells_to_update = [
{"row_no": 1, "column_name": "query", "value": "Updated query text"},
{"row_no": 2, "column_name": "response", "value": "New model response"}
]
# Update the cells in the dataset
try:
result = Dataset.update_cells("your-dataset-id", cells_to_update)
print(f"Successfully updated cells: {result}")
except Exception as e:
print(f"Error updating cells: {e}")
Parameters
dataset_id
(str): The ID of the dataset to update cells in.
cells
(List[Dict]): A list of cells to update, where each cell is a dictionary containing:
row_no
(int): The row number (1-based indexing) of the cell to update.
column_name
(str): The name of the column containing the cell to update.
value
(Any): The new value for the specified cell.
Return Value
The method returns a dictionary with the API response, typically containing a success message.
Error Handling
If the API call fails, a CustomException
is raised with details about the error.
Direct API Calls
You can also update cells by making a direct API call:
curl --location --request PUT 'http://api.athina.ai/api/v1/dataset_v2/<DATASET_ID>/cells' \
--header 'athina-api-key: <ATHINA_API_KEY>' \
--header 'Content-Type: application/json' \
--data '{
"cells": [
{
"row_no": 1,
"column_name": "query",
"value": "Hello World"
},
{
"row_no": 3,
"column_name": "response",
"value": "Greetings from Athina!"
}
]
}'
- Method: PUT
- URL:
http://api.athina.ai/api/v1/dataset_v2/<DATASET_ID>/cells
- Headers:
athina-api-key
: Your Athina API key
Content-Type
: application/json
- Body:
{
"cells": [
{
"row_no": <row_number>,
"column_name": "<column_name>",
"value": <new_value>
},
...
]
}
If the updates are successful, you will receive a response similar to:
{
"status": "success",
"data": {
"message": "Cells updated successfully"
}
}
Responses are generated using AI and may contain mistakes.