* feat: add DataFrame.ai.forecast() support

* test

* fix

* constructor

* update

* comments

* fix

* feat: add describe() method to Series

* fix test

* update docs

* fix tests

* feat: add required param 'engine' to multimodal functions

* add missing engine to exif test

* 🦉 Updates from OwlBot post-processor

See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md

---------

Co-authored-by: Tim Sweña (Swast) <[email protected]>
Co-authored-by: Owl Bot <gcf-owl-bot[bot]@users.noreply.github.com>
Release-As: 2.8.0

…interactive display of the results (#1820)

* add anywidget as extra python package

* Add anywidget to bigframes

* add the first testcase

* fix mypy error

* Show first page of results (to_pandas_batches()) is done

* add the testcase

* Add more testcase

* change a import

* change anywidget mode for plain text

* fix noxfile

* add anywidget for docx

* ignore missing import

* fix doctest

* add unittest

* add notebook test

* fix unit-10.12

* change testcase

* make anywidget addtional

* remove anywidget_mode.ipynb in notebook session due to deferred mode

* fix typo

* fix failed testcase

* chore: move `_read_gbq_colab` to top-level `bigframes.pandas`

* 🦉 Updates from OwlBot post-processor

See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md

---------

Co-authored-by: Owl Bot <gcf-owl-bot[bot]@users.noreply.github.com>

…lient data download (#1829)

* feat: Add maximum_rows_downloaded option to limit client data download

This commit introduces a new compute option `bigframes.pandas.options.compute.maximum_rows_downloaded`
that allows you to set a limit on the maximum number of rows that can be downloaded to a client machine.

When this option is set and a data-downloading operation (e.g., `to_pandas()`, `to_pandas_batches()`)
attempts to download more rows than the configured limit, a `bigframes.exceptions.MaximumRowsDownloadedExceeded`
exception is raised.

This feature helps prevent Out-Of-Memory (OOM) errors in shared execution environments
by providing a mechanism to control the amount of data downloaded to the client.

The limit is checked in both `DirectGbqExecutor` and `BigQueryCachingExecutor`.
Unit tests have been added to verify the functionality, including scenarios
where the limit is not set, set but not exceeded, and set and exceeded for various
DataFrame operations.
Documentation has been updated by ensuring the docstring for the new option
in `ComputeOptions` is comprehensive for automatic generation.

* refactor: Use a shared helper function for row limit checks

This commit refactors the row limit check logic in `DirectGbqExecutor`
and `BigQueryCachingExecutor` to use a new shared helper function
`check_row_limit` located in `bigframes.session.utils`.

This change reduces code duplication and improves maintainability.
The functionality remains the same as before the refactoring.

* address review comments

* fix imports

* fix docs

* rename to maximum_result_rows

* cleanup

* restore deleted test

---------

Co-authored-by: google-labs-jules[bot] <161369871+google-labs-jules[bot]@users.noreply.github.com>

Co-authored-by: release-please[bot] <55107282+release-please[bot]@users.noreply.github.com>