Add test for date ranges

evidentlyai / evidently

Evidently is an open-source ML and LLM observability framework. Evaluate, test, and monitor any AI-powered system or data pipeline. From tabular data to Gen AI. 100+ metrics.

https://www.evidentlyai.com/evidently-oss

Apache License 2.0

5.3k stars 594 forks source link

Add test for date ranges #462

Open elenasamuylova opened 1 year ago

elenasamuylova commented 1 year ago

Test Group: Data Quality

Feature request: to add the ability to test that values in a DateTime column are within a specific range.

Source: Discord conversation https://discord.com/channels/815991896916361276/815992391076806657/threads/1044973406929571860

AntonisCSt commented 1 year ago

Hi, to address this, would it be possible to adjust the following classes:

class TestValueRange from \tests\data_quality_tests.py
and any other inherited class such as ColumnValueRangeMetric from metrics\data_quality\column_value_list_metric.py

Adjust them to accept Timestamp types (right and left) and not only Numeric.

The column.between() pandas method (that is used in ColumnValueRangeMetric ) can be used with Timestamps.

What do you think?

emeli-dral commented 1 year ago

Hi @AntonisCSt , This sounds like a good approach!

I'm very sorry for the late reply. The things is that we worked on MetricResult refactoring and we didn't want to update any metric before we finish. We finally transitioned MetricResult classes from DataClass to Pydantic in the latest release, and now ready to update Metrics.

Would you consider making a contribution?

AntonisCSt commented 1 year ago

Hi @emeli-dral, Thank you for the update! I would be definitely happy to make a contribution.

So now with the new approach are there any specific guidelines or steps I should follow to get started? If not, I can move on and update the TestValueRange and ColumnValueRangeMetric classes to accept Timestamp types.