Release 0.8.0

[nnansters]

Hi all,

Niels from engineering here to tell you about the 0.8.0 release!

Installing / upgrading

You can get this latest version by using pip:

pip install -U nannyml

Or conda:

conda install -c conda-forge nannyml

What's new?

More univariate drift methods!

Our data science and research teams have been investigating and benchmarking methods for detecting univariate data drift.
Two methods made the cut and are now available in the library: the Wasserstein distance and the L-Infinity distance.
You can find out more about them in the univariate drift methods documentation.

You can use them with the UnivariateDriftCalculator as follows:

import nannyml as nml
from IPython.display import display

reference_df = nml.load_synthetic_binary_classification_dataset()[0]
analysis_df = nml.load_synthetic_binary_classification_dataset()[1]
display(reference_df.head())

column_names = ['distance_from_office', 'salary_range', 'gas_price_per_litre', 'public_transportation_cost', 
                                 'wfh_prev_workday', 'workday', 'tenure', 'y_pred_proba', 'y_pred']
calc = nml.UnivariateDriftCalculator(
    column_names=column_names,
    timestamp_column_name='timestamp',
    continuous_methods=['wasserstein', 'jensen_shannon'],
    categorical_methods=['l_infinity', 'jensen_shannon'],
)

calc.fit(reference_df)
results = calc.calculate(analysis_df)
display(results.filter(period='analysis', column_names=['distance_from_office']).to_df())

Usage logging

Given that we've not heard any outcry against introducing usage logging after our announcement blog, we've decided to move its launch forward to better match our marketing activities.

You can find out all you want to know about the why, what, how and how to disable usage logging on the usage logging documentation page.

What's changed?

• Some additions and fixes to the documentation
• An update of mypy led to some new issues to resolve

What's next?

We have some exciting, foundational updates coming very, very soon. Stay tuned!