Managing Datasets: Delete or Update Methods for client.datasets

·Oct 02, 2025 08:29 PM·

There seems to be a limited number of methods on the client.datasets object. In particular, how do I delete a dataset, or update it with a new version? Currently it seems I am forced to delete it from the UI?

👀1

7 comments

· Sorted by Oldest

RunLLM
APP
·
·
You can delete a dataset programmatically using the Python client or REST API; the `client.datasets` object supports deletion via the `delete` method, which calls the `/v1/datasets/{id}` endpoint (<https://arize.com/docs/phoenix/sdk-api-reference/rest-api/datasets|API docs>, <https://arize-ai.github.io/phoenix/interfaces/_arizeai_phoenix-client.__generated___api_v1.paths.html#v1datasetsid|TypeScript API>). For updating a dataset with new examples, use `add_examples_to_dataset`, which appends data and creates a new version—however, you cannot overwrite or replace the entire dataset in place; full upsert or replacement is not currently supported (<https://github.com/Arize-ai/phoenix/blob/main/packages/phoenix-client/docs/source/index.md|client docs>, <https://github.com/Arize-ai/phoenix/issues/9303|issue discussion>). If you need to update the dataset content entirely, the current workaround is to delete the dataset and recreate it, or use versioning by appending new examples. Direct in-place editing or overwriting of all examples is not available via the SDK or API as of now (<https://github.com/Arize-ai/phoenix/issues/9303|issue #9303>, <https://github.com/Arize-ai/phoenix/blob/main/.cursor/rules/client-migration.mdc|client migration guide>). Would you like more detail or code examples for these operations? Sources: - <https://arize.com/docs/phoenix/sdk-api-reference/rest-api/datasets|API docs> - <https://arize-ai.github.io/phoenix/interfaces/_arizeai_phoenix-client.__generated___api_v1.paths.html#v1datasetsid|TypeScript API> - <https://github.com/Arize-ai/phoenix/blob/main/packages/phoenix-client/docs/source/index.md|client docs> - <https://github.com/Arize-ai/phoenix/issues/9303|issue discussion> - <https://github.com/Arize-ai/phoenix/blob/main/.cursor/rules/client-migration.mdc|client migration guide>
💡 Hint: Mention RunLLM in the thread for followups.
Sanjana
External
·
Hi Rasmus B. - the bot is correct here. This is the more exhaustive list of what you can do with datasets using the REST API (including delete). https://arize.com/docs/phoenix/sdk-api-reference/rest-api/api-reference/datasets
Sanjana
External
·
When you add examples to the data (update) it will be versioned
Rasmus B.
·
·
thanks 🔒[private user] feel like I searched the python API and did not see the delete method, will check again. Otherwise I will try REST. Is there a shortcut for sending REST calls through the client with the correct headers etc?
Rasmus B.
·
·
RunLLM feel like I searched the python API and did not see the delete method. Is there otherwise a shortcut for sending REST calls through the client with the correct headers etc?
RunLLM
APP
·
·
The Python client does not expose a direct `delete` method on `client.datasets` in all versions, but you can use the underlying HTTP session of the client to send REST calls with the correct headers and authentication automatically handled (<https://github.com/Arize-ai/phoenix/blob/main/packages/phoenix-client/docs/source/index.md|client docs>). For example, you can use `client._client.delete("/v1/datasets/{id}")` to send a DELETE request with proper headers. Would you like a code example for this approach? Sources: - <https://github.com/Arize-ai/phoenix/blob/main/packages/phoenix-client/docs/source/index.md|client docs>
💡 Hint: Mention RunLLM in the thread for followups.
Sanjana
External
·
Hi Rasmus B.! Sorry for the confusion in my wording - you are right there is no delete method in teh python API. REST is the way to go like the bot described
👍1

RunLLM
APP
·
·
You can delete a dataset programmatically using the Python client or REST API; the `client.datasets` object supports deletion via the `delete` method, which calls the `/v1/datasets/{id}` endpoint (<https://arize.com/docs/phoenix/sdk-api-reference/rest-api/datasets|API docs>, <https://arize-ai.github.io/phoenix/interfaces/_arizeai_phoenix-client.__generated___api_v1.paths.html#v1datasetsid|TypeScript API>). For updating a dataset with new examples, use `add_examples_to_dataset`, which appends data and creates a new version—however, you cannot overwrite or replace the entire dataset in place; full upsert or replacement is not currently supported (<https://github.com/Arize-ai/phoenix/blob/main/packages/phoenix-client/docs/source/index.md|client docs>, <https://github.com/Arize-ai/phoenix/issues/9303|issue discussion>). If you need to update the dataset content entirely, the current workaround is to delete the dataset and recreate it, or use versioning by appending new examples. Direct in-place editing or overwriting of all examples is not available via the SDK or API as of now (<https://github.com/Arize-ai/phoenix/issues/9303|issue #9303>, <https://github.com/Arize-ai/phoenix/blob/main/.cursor/rules/client-migration.mdc|client migration guide>). Would you like more detail or code examples for these operations? Sources: - <https://arize.com/docs/phoenix/sdk-api-reference/rest-api/datasets|API docs> - <https://arize-ai.github.io/phoenix/interfaces/_arizeai_phoenix-client.__generated___api_v1.paths.html#v1datasetsid|TypeScript API> - <https://github.com/Arize-ai/phoenix/blob/main/packages/phoenix-client/docs/source/index.md|client docs> - <https://github.com/Arize-ai/phoenix/issues/9303|issue discussion> - <https://github.com/Arize-ai/phoenix/blob/main/.cursor/rules/client-migration.mdc|client migration guide>
💡 Hint: Mention RunLLM in the thread for followups.
Sanjana
External
·
Hi Rasmus B. - the bot is correct here. This is the more exhaustive list of what you can do with datasets using the REST API (including delete). https://arize.com/docs/phoenix/sdk-api-reference/rest-api/api-reference/datasets
Sanjana
External
·
When you add examples to the data (update) it will be versioned
Rasmus B.
·
·
thanks 🔒[private user] feel like I searched the python API and did not see the delete method, will check again. Otherwise I will try REST. Is there a shortcut for sending REST calls through the client with the correct headers etc?
Rasmus B.
·
·
RunLLM feel like I searched the python API and did not see the delete method. Is there otherwise a shortcut for sending REST calls through the client with the correct headers etc?
RunLLM
APP
·
·
The Python client does not expose a direct `delete` method on `client.datasets` in all versions, but you can use the underlying HTTP session of the client to send REST calls with the correct headers and authentication automatically handled (<https://github.com/Arize-ai/phoenix/blob/main/packages/phoenix-client/docs/source/index.md|client docs>). For example, you can use `client._client.delete("/v1/datasets/{id}")` to send a DELETE request with proper headers. Would you like a code example for this approach? Sources: - <https://github.com/Arize-ai/phoenix/blob/main/packages/phoenix-client/docs/source/index.md|client docs>
💡 Hint: Mention RunLLM in the thread for followups.
Sanjana
External
·
Hi Rasmus B.! Sorry for the confusion in my wording - you are right there is no delete method in teh python API. REST is the way to go like the bot described
👍1