Quickstart

What is NannyML?

NannyML detects silent model failure, estimates performance of ML models after deployment before target data become available and robustly detects data drift potentially responsible for the failure.

Installing NannyML

From the shell of your python environment type:

$ pip install nannyml

Content of quickstart

This quickstart presents core functionalities of NannyML on an example of binary classification model that predicts whether an employee will work from home the next day or not. First, the whole code is shown so experiment-first users can get their hands dirty right away. This is followed by detailed walk through so read-first minds can get familiar with the flow before coding. The synthetic dataset used is already preprocessed - it is merged with model predictions and ready to be directly used by NannyML. In order to find out how to prepare your own dataset check these tutorials. This is also a good place to go to get detailed guides on main concepts and functionalities. If you want to know what is implemented under the hood - visit how it works. Finally, if you just look for examples on other datasets or ML problems look at examples.

Just the code

Feel free to copy-paste the code below and experiment with the results yourself. If you want us to guide you go to the Walkthrough.

>>> import pandas as pd
>>> import nannyml as nml
>>> from IPython.display import display
>>> reference, analysis, analysis_target = nml.load_synthetic_binary_classification_dataset()
>>> display(analysis.head())
>>> display(reference.head())

>>> metadata = nml.extract_metadata(data = reference, model_name='wfh_predictor', model_type='classification_binary', exclude_columns=['identifier'])
>>> metadata.target_column_name = 'work_home_actual'
>>> data = pd.concat([reference, analysis], ignore_index=True)
>>> chunk_size = 5000

>>> # initialize, fit estimator and estimate
>>> estimator = nml.CBPE(model_metadata=metadata, chunk_size=chunk_size, metrics=['roc_auc'])
>>> estimator = estimator.fit(reference)
>>> estimated_performance = estimator.estimate(data=data)
>>> # show results
>>> figure = estimated_performance.plot(kind='performance', metric='roc_auc')
>>> figure.show()

>>> # Let's initialize the object that will perform the Univariate Drift calculations
>>> univariate_calculator = nml.UnivariateStatisticalDriftCalculator(model_metadata=metadata, chunk_size=chunk_size)
>>> univariate_calculator = univariate_calculator.fit(reference_data=reference)
>>> univariate_results = univariate_calculator.calculate(data=data)
>>> # let's plot drift results for all model inputs
>>> for feature in metadata.features:
...     figure = univariate_results.plot(kind='feature_drift', metric='statistic', feature_label=feature.label)
...     figure.show()

>>> ranker = nml.Ranker.by('alert_count')
>>> ranked_features = ranker.rank(univariate_results, model_metadata=metadata, only_drifting = False)
>>> display(ranked_features)

>>> figure = univariate_results.plot(kind='prediction_drift', metric='statistic')
>>> figure.show()

>>> # Let's initialize the object that will perform Data Reconstruction with PCA
>>> rcerror_calculator = nml.DataReconstructionDriftCalculator(model_metadata=metadata, chunk_size=chunk_size).fit(reference_data=reference)
>>> # let's see Reconstruction error statistics for all available data
>>> rcerror_results = rcerror_calculator.calculate(data=data)
>>> figure = rcerror_results.plot(kind='drift')
>>> figure.show()

Walk through the quickstart

Let’s start with loading the synthetic dataset included in the library:

>>> import pandas as pd
>>> import nannyml as nml
>>> reference, analysis, analysis_target = nml.load_synthetic_binary_classification_dataset()
>>> reference.head()

	distance_from_office	salary_range	gas_price_per_litre	public_transportation_cost	wfh_prev_workday	workday	tenure	identifier	work_home_actual	timestamp	y_pred_proba	partition	y_pred
0	5.96225	40K - 60K €	2.11948	8.56806	False	Friday	0.212653	0	1	2014-05-09 22:27:20	0.99	reference	1
1	0.535872	40K - 60K €	2.3572	5.42538	True	Tuesday	4.92755	1	0	2014-05-09 22:59:32	0.07	reference	0
2	1.96952	40K - 60K €	2.36685	8.24716	False	Monday	0.520817	2	1	2014-05-09 23:48:25	1	reference	1
3	2.53041	20K - 20K €	2.31872	7.94425	False	Tuesday	0.453649	3	1	2014-05-10 01:12:09	0.98	reference	1
4	2.25364	60K+ €	2.22127	8.88448	True	Thursday	5.69526	4	1	2014-05-10 02:21:34	0.99	reference	1

The synthetic dataset provided contains inputs and predictions of a binary classification model that predicts whether an employee will work from home the next workday or not. The probability of the employee working from home is included in the y_pred_proba column, while the prediction is in y_pred column. The model inputs are distance_from_office, salary_range, gas_price_per_litre, public_transportation_cost, wfh_prev_workday, workday and tenure. identifier is the Identifier column and timestamp is the Timestamp column.

The next step is to have NannyML extract model metadata from the dataset and make a choice about the way we will split our data into Data Chunks.

>>> metadata = nml.extract_metadata(data = reference, model_name='wfh_predictor', model_type='classification_binary', exclude_columns=['identifier'])
>>> metadata.target_column_name = 'work_home_actual'
>>> data = pd.concat([reference, analysis], ignore_index=True)
>>> chunk_size = 5000

The data are already split into a reference and an analysis periods. NannyML uses the reference period to establish a baseline for expected model performance. The analysis period is where we estimate or monitor performance, as well as detect data drift. For more information about periods check Data Periods. The key thing is that the analysis period doesn’t need to contain the Target data. Therefore in our example it is provided as a separate object.

>>> analysis.head()

	distance_from_office	salary_range	gas_price_per_litre	public_transportation_cost	wfh_prev_workday	workday	tenure	identifier	timestamp	y_pred_proba	partition	y_pred
0	0.527691	0 - 20K €	1.8	8.96072	False	Tuesday	4.22463	50000	2017-08-31 04:20:00	0.99	analysis	1
1	8.48513	20K - 20K €	2.22207	8.76879	False	Friday	4.9631	50001	2017-08-31 05:16:16	0.98	analysis	1
2	2.07388	40K - 60K €	2.31008	8.64998	True	Friday	4.58895	50002	2017-08-31 05:56:44	0.98	analysis	1
3	0.118456	20K - 20K €	2.17144	8.85542	False	Tuesday	4.71101	50003	2017-08-31 06:10:17	0.97	analysis	1
4	4.7867	0 - 20K €	2.36854	8.39497	False	Monday	0.906738	50004	2017-08-31 06:29:38	0.92	analysis	1

Estimating Performance without Targets

NannyML can estimate the performance on a machine learning model in production without access to its Target. For more details on how to use performance estimation see here, while for more details on how it works see here.

>>> # initialize, fit estimator and estimate
>>> estimator = nml.CBPE(model_metadata=metadata, chunk_size=chunk_size, metrics=['roc_auc'])
>>> estimator = estimator.fit(reference)
>>> estimated_performance = estimator.estimate(data=data)
>>> # show results
>>> figure = estimated_performance.plot(kind='performance', metric='roc_auc')
>>> figure.show()

The results indicate that the model’s performance is likely to be negatively impacted from the second half of 2019 onwards.

Detecting Data Drift

NannyML allows for further investigation into potential performance issues with its data drift detection functionality. See Detecting Data Drift for more details.

>>> # Let's initialize the object that will perform the Univariate Drift calculations
>>> univariate_calculator = nml.UnivariateStatisticalDriftCalculator(model_metadata=metadata, chunk_size=chunk_size)
>>> univariate_calculator = univariate_calculator.fit(reference_data=reference)
>>> univariate_results = univariate_calculator.calculate(data=data)
>>> # let's plot drift results for all model inputs
>>> for feature in metadata.features:
...     figure = univariate_results.plot(kind='feature_drift', metric='statistic', feature_label=feature.label)
...     figure.show()

When there are a lot of drifted features, NannyML can also rank them by the number of alerts they have raised:

>>> ranker = nml.Ranker.by('alert_count')
>>> ranked_features = ranker.rank(univariate_results, model_metadata=metadata, only_drifting = False)
>>> ranked_features

	feature	number_of_alerts	rank
0	wfh_prev_workday	5	1
1	salary_range	5	2
2	distance_from_office	5	3
3	public_transportation_cost	5	4
4	tenure	2	5
5	workday	0	6
6	gas_price_per_litre	0	7

Drift in the model outputs can be also visualized:

>>> figure = univariate_results.plot(kind='prediction_drift', metric='statistic')
>>> figure.show()

More complex data drift cases can get detected by Data Reconstruction with PCA. For more information see Data Reconstruction with PCA.

>>> # Let's initialize the object that will perform Data Reconstruction with PCA
>>> rcerror_calculator = nml.DataReconstructionDriftCalculator(model_metadata=metadata, chunk_size=chunk_size).fit(reference_data=reference)
>>> # let's see Reconstruction error statistics for all available data
>>> rcerror_results = rcerror_calculator.calculate(data=data)
>>> figure = rcerror_results.plot(kind='drift')
>>> figure.show()

Putting everything together, we see that 4 features exhibit data drift from late 2019 onwards. They are distance_from_office, salary_range, public_transportation_cost, wfh_prev_workday. This drift is responsible for the potential negative impact in performance that we observed.

Insights and Follow Ups

With NannyML we were able to estimate performance in the absence of ground truth. The estimation has shown potential drop in ROC AUC in the second half of the analysis period. Univariate and multivariate data drift detection algorithms have identified data drift in this period, potentially justifying the drop. This could be further investigated by analyzing changes of distributions of the input variables. Check tutorials on data drift to find out how to plot distributions with NannyML.