Data drift detection in MLOps - Interactive Code Practice

Practice - 5 Tasks

1fill in blank

easy

Complete the code to import the library used for data drift detection.

MLOps

from evidently.[1] import Report

Drag options to blanks, or click blank then click option'

Adashboard

Bdrift

Cmetrics

Dreport

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'dashboard' which is for visual display, not report generation.

Using 'metrics' which is for individual measurements, not full reports.

The report module from Evidently is used to create data drift detection reports.

2fill in blank

medium

Complete the code to create a data drift report object.

MLOps

data_drift_report = [1](metrics=[DataDriftTable()])

Drag options to blanks, or click blank then click option'

ADashboard

BAnalyzer

CReport

DProfile

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'Dashboard' which is for interactive visualizations.

Using 'Profile' which is a different class for profiling data.

The Report class is used to create a report object that can include data drift metrics.

3fill in blank

hard

Fix the error in the code to run the data drift report on reference and current data.

MLOps

data_drift_report.[1](reference_data=ref_data, current_data=curr_data)

Drag options to blanks, or click blank then click option'

Arun

Bexecute

Cstart

Dlaunch

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'execute' which is not a method of the report object.

Using 'start' or 'launch' which do not exist in this context.

The run method executes the report calculation on the given datasets.

4fill in blank

hard

Fill both blanks to save the data drift report as an HTML file.

MLOps

with open('drift_report.[1]', '[2]') as f:
    f.write(data_drift_report.html())

Drag options to blanks, or click blank then click option'

Ahtml

Bw

Ctxt

Dr

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'txt' extension which is not suitable for HTML content.

Using 'r' mode which is for reading, not writing.

The report is saved as an HTML file, so the extension is html. The file mode to write is w.

5fill in blank

hard

Fill all three blanks to create a dictionary of drift metrics filtered by threshold.

MLOps

drift_metrics = {k: v for k, v in data_drift_report.[1]().items() if v [2] [3]

Drag options to blanks, or click blank then click option'

Aas_dict

B>

C0.1

Drun

Attempts:

3 left

💡 Hint

Common Mistakes

Using 'run' instead of 'as_dict' which is not a method to get dictionary data.

Using '<' instead of '>' which would filter the wrong values.

The as_dict() method returns the report data as a dictionary. The filter keeps metrics greater than 0.1.