Add dashboard panels

How to add and configure monitoring panels.

New dashboards are empty by default. You must define the dashboard composition in the code. You can choose which values or test results to display and select from several monitoring panel types.

Code example

Refer to the QuickStart Tutorial for a complete Python script with multiple monitoring panels.

pageSelf-host ML Monitoring

You can also explore live demo dashboards and the corresponding source code.

How it works

Evidently snapshots contain multiple measurements. For example, when you log the DataDriftTable() Metric in a snapshot, it will contain the dataset drift summary, similar to this:

'number_of_columns': 15,
'number_of_drifted_columns': 5,
'share_of_drifted_columns': 0.3333333333333333,
'dataset_drift': False,

It will also contain data on individual column drift. Here is a partial example:

'column_name': 'age',
'column_type': 'num',
'stattest_name': 'Wasserstein distance (normed)',
'stattest_threshold': 0.1,
'drift_score': 0.18534692319042428,
'drift_detected': True,

The same logic applies to other Metrics and Tests.

You can visualize any measurement captured in the snapshots over time. To do that, you must add a panel to a monitoring dashboard of a specific project and specify the value you'd like to plot.

For example, if you logged the DataDriftTable() metric, you may later choose to plot measurements like:

• share_of_drifted_columns

• number_of_drifted_columns

• drift_score for a specific column.

All these measurements will be available as MetricResults inside the snapshot.

To create a monitoring panel, you will also need to specify other parameters, such as panel type, width, title and legend. This docs section explains how.

Add panel

To add a new panel to an existing dashboard, use the add_panel() method.

Example. To add a new Counter panel showing the share of drifting columns:

project.dashboard.add_panel(
    DashboardPanelCounter(
        title="Share of Drifted Features",
        filter=ReportFilter(metadata_values={}, tag_values=[]),
        value=PanelValue(
            metric_id="DatasetDriftMetric",
            field_path="share_of_drifted_columns",
            legend="share",
         ),
         text="share",
         agg=CounterAgg.LAST,
         size=WidgetSize.HALF,
    )
)

Note: project.dashboard is an exemplar of the DashboardConfig class.

You can add multiple panels to a project dashboard. They will appear in the order listed in the project.

Panel types

You can choose between the following panel types.

Panel Type	Example
Metric counter (DashboardPanelCounter) Shows a measurement (a MetricResult from a Report or Test Suite) with supporting text. Also works as a text-only panel, e.g. a title panel.
Metric plot (DashboardPanelPlot) Shows a measurement (a MetricResult from a Report or Test Suite) as a line plot, bar plot, scatter plot or histogram.	See below.
Line plot (PlotType.LINE)
Scatter plot (PlotType.SCATTER)
Bar plot (PlotType.BAR)
Histogram (PlotType.HISTOGRAM)
Test counter (DashboardPanelTestSuiteCounter) Shows a counter of failed or passed tests.
Test plot (DashboardPanelTestSuite) Shows failed and passed tests in time. You can choose a detailed view (with individual test results) or aggregate (with the total number of tests).	See below.
Detailed plot (TestSuitePanelType.DETAILED)
Aggregate plot (TestSuitePanelType.AGGREGATE)

Panel parameters

Class DashboardPanel

This is a base class. The parameters below apply to all panel types. There are also panel-specific parameters explained in the following sections.

Parameter	Description
title: str Usage: title="My New Panel”	The name of the panel. It will be visible at the header of a panel on a dashboard.
filter: ReportFilter metadata_values: Dict[str, str] tag_values: List[str] include_test_suites=False Usage: filter=ReportFilter(metadata_values={}, tag_values=[], include_test_suites=True)	Filters help specify a subset of snapshots from which to display the values. To use filtering, you must add metadata or tags when logging Reports or Test Suites. (See docs). You can then reference these metadata_values or tag_values when creating a panel. To include data from Test Suites, you must set the include_test_suites parameter as True (default: False).
size: WidgetSize = WidgetSize.FULL Available: WidgetSize.FULL, WidgetSize.HALF Usage: size=WidgetSize.FULL	Sets the size of the panel as half-width or full-sized (Default).

DashboardPanelCounter

DashboardPanelCounter helps add metric counters or text panels. You can pull metric values from both Reports and Test Suites.

Example 1. To create a panel with the dashboard title only:

project.dashboard.add_panel(
    DashboardPanelCounter(
        filter=ReportFilter(metadata_values={}, tag_values=[]),
        agg=CounterAgg.NONE,
        title="Bike Rental Demand Forecast",
        )
    )

Example 2. To create a panel that sums up measurements (number of rows) over time.

project.dashboard.add_panel(
    DashboardPanelCounter(
        title="Model Calls",
        filter=ReportFilter(metadata_values={}, tag_values=[]),
        value=PanelValue(
            metric_id="DatasetSummaryMetric",
            field_path=DatasetSummaryMetric.fields.current.number_of_rows,
            legend="count",
        ),
        text="count",
        agg=CounterAgg.SUM,
        size=WidgetSize.HALF,
    )
)

Parameter	Description
value: Optional[PanelValue] = None	The value (MetricResult) to show in the Counter. You can create a simple text panel if you do not pass the Value. See the section below on Panel Values for more examples.
text: Optional[str] = None	Supporting text to be shown on the Counter.
agg: CounterAgg Available: SUM, LAST, NONE	Data aggregation options: SUM - sums the values from different snapshots over time considering the applied filters (if any). LAST - shows the last available value NONE - to be used for text panels

DashboardPanelPlot

DashboardPanelPlot allows creating scatter, bar, line, and histogram plots with metric values. You can pull metric values from both Reports and Test Suites.

Example. To plot MAPE over time in a line plot.

project.dashboard.add_panel(
    DashboardPanelPlot(
        title="MAPE",
        filter=ReportFilter(metadata_values={}, tag_values=[]),
        values=[
        PanelValue(
            metric_id="RegressionQualityMetric",
            field_path=metrics.RegressionQualityMetric.fields.current.mean_abs_perc_error,
            legend="MAPE",
        ),
    ],
    plot_type=PlotType.LINE,
    size=WidgetSize.HALF,
    )
)

Parameter	Description
values: List[PanelValue]	You must pass at least one value (MetricResult). You can also pass multiple values as a list. They will appear together: for example, as separate lines on a Line plot, bars on a Bar Chart, or points on a Scatter Plot. If you use a Histogram, the values will be aggregated. See the section below on Panel Values for more examples.
plot_type: PlotType Available: SCATTER, BAR, LINE, HISTOGRAM	Specifies the plot type.

DashboardPanelTestSuiteCounter

DashboardPanelTestSuiteCounter displays a counter of failed and passed tests. It applies to Test Suites only.

Example 1. To display the result of the last test.

project.dashboard.add_panel(
    DashboardPanelTestSuiteCounter(
        title="Success of last",
        agg=CounterAgg.LAST
    )
)

Example 2. To display the number of failed tests and errors in the test results for a specific column.

project.dashboard.add_panel(
    DashboardPanelTestSuiteCounter(
        title="Success of 1",
        test_filters=[TestFilter(test_id="TestNumberOfUniqueValues", test_args={"column_name.name": "1"})],
        statuses=[TestStatus.ERROR, TestStatus.FAIL]
    )
)

See applicable parameters in the following section.

DashboardPanelTestSuite

DashboardPanelTestSuite displays the results of failed and passed tests over time. It applies to Test Suites only.

Example 1. To show the results of all individual tests over time, with daily level aggregation.

project.dashboard.add_panel(
    DashboardPanelTestSuite(
        title="All tests: detailed",
        filter=ReportFilter(metadata_values={}, tag_values=[], include_test_suites=True),
        size=WidgetSize.HALF,
        panel_type=TestSuitePanelType.DETAILED,
        time_agg="1D",
    )
)

Example 2. To show the results of individual tests for specific columns, with daily aggregation.

project.dashboard.add_panel(
    DashboardPanelTestSuite(
        title="Column Drift tests for key features: detailed",
        test_filters=[
            TestFilter(test_id="TestColumnDrift", test_args={"column_name.name": "hours-per-week"}),
            TestFilter(test_id="TestColumnDrift", test_args={"column_name.name": "capital-gain"}),
        ],
        filter=ReportFilter(metadata_values={}, tag_values=[], include_test_suites=True),
        size=WidgetSize.HALF,
        panel_type=TestSuitePanelType.DETAILED,
        time_agg="1D",
    )
)

Example 3. To show the number of passed and failed tests, with daily level aggregation.

project.dashboard.add_panel(
    DashboardPanelTestSuite(
        title="All tests: aggregated",
        filter=ReportFilter(metadata_values={}, tag_values=[], include_test_suites=True),
        size=WidgetSize.HALF,
        time_agg="1D",
    )
)

Parameter	Description
test_filters: List[TestFilter]=[] test_id: test_id test_arg: List[str] Usage: test_filters=[TestFilter(test_id="TestNumberOfUniqueValues", test_args={"column_name.name": "1"})]	Filters that help include the results only for specific Tests and/or columns. If not specified, all logged tests will be considered.
statuses: List[statuses] Available: TestStatus.ERROR, TestStatus.FAIL, TestStatus.SUCCESS, TestStatus.WARNING, TestStatus.SKIPPED Usage: statuses=[TestStatus.ERROR, TestStatus.FAIL]	Filters that help include only the test results with a specific status. If not specified, tests with any status will be considered.
agg: CounterAgg Available: SUM, LAST Usage: agg=CounterAgg.LAST	Applies to the DashboardPanelTestSuiteCounter. Data aggregation options: SUM - sums the test results from multiple snapshots, considering the applied filters (if any). LAST - shows the last available test result.
panel_type=TestSuitePanelType Available: TestSuitePanelType.DETAILED TestSuitePanelType.AGGREGATE Usage: panel_type=TestSuitePanelType.DETAILED	Applies to the DashboardPanelTestSuite. You can use it to define the type of the plot: Detailed (includes the individual test results) Aggregate (default, includes only the total number of passed/failed tests)
time_agg: Optional[str] = None Available: 1H, 1D, 1W, 1M (see period aliases) Usage: time_agg="1D"	Applies to the DashboardPanelTestSuite. Aggregation options to group the test results within a defined time period on a plot.

Panel value

To add a numerical measurement to the plot, you must pass the PanelValue. For example, you can display the number of drifting features, the share of empty columns, mean error, etc.

Parameters. To define which values to show on a specific panel, you must specify:

Parameter	Description
metric_id	A metric ID that corresponds to the Evidently Metric logged inside the snapshots. You must specify the metric_id even if you use Test Suites.
field_path	The path that corresponds to the specific MetricResult computed as part of this Metric or Test. You can pass either a complete field path or a "field_name".
metric_args (optional)	Additional arguments that specify the metrics parameters. This is applicable when multiple instances of the same metric are logged in a snapshot. For example: column name, text descriptor, drift detection method used, etc.
legend (optional)	The legend that will be visible in the plot.

Example 1. To include the share_of_drifted_columns MetricResult, available inside the DatasetDriftMetric():

value=PanelValue(
    metric_id="DatasetDriftMetric",
    field_path="share_of_drifted_columns",
    legend="share",
)

In this example, you pass the exact name of the field.

Example 2. To include the current.share_of_missing_values available inside the DatasetMissingValueMetric():

value=PanelValue(
    metric_id="DatasetMissingValuesMetric",
    field_path=DatasetMissingValuesMetric.fields.current.share_of_missing_values,
    legend="Missing Values Share",
)

In this example, you pass the complete field path inside the Metric.

Note. You must always reference a metric_id, even if you used a Preset. For example, if you used a DataDriftPreset(), you can reference either of the Metrics it contains (DataDriftTable() or DatasetDriftMetric()). You can verify the Metrics included in each Preset in the reference table.

Example 3. To display the mean values of target and prediction over time in a line plot.

values=[
    PanelValue(
        metric_id="ColumnSummaryMetric",
        field_path="current_characteristics.mean",
        metric_args={"column_name.name": "cnt"},
        legend="Target (daily mean)",
    ),
    PanelValue(
        metric_id="ColumnSummaryMetric",
        field_path="current_characteristics.mean",
        metric_args={"column_name.name": "prediction"},
        legend="Prediction (daily mean)",
    ),
]

In this example, you pass additional metric arguments to specify the column names.

Example 4. To specify the drift detection method (when results for multiple methods logged inside a snapshot) using metric_args.

values=[
    PanelValue(
        metric_id="EmbeddingsDriftMetric", 
        metric_args={"drift_method.dist": "euclidean"}
    )
}

Example 5. To specify the text descriptor using metric_args.

values=[
    PanelValue(
        metric_id="ColumnSummaryMetric",
        metric_args={"column_name": descriptors.OOV(display_name="OOV").for_column("Review_Text")},
        field_path="current_characteristics.mean",
        legend="OOV % (mean)",
    ),
]

How to find the field path?

Option 1. Use autocomplete.

You can use autocomplete in interactive Python environments (like Jupyter notebook or Colab) to see available fields inside a specific Metric. They appear as you start typing the .fields. path for a specific Metric.

Note: some types of values (e.g. mean, sum, max, min) will not be visible using this method. This is because they match the names of the standard Python fields.

Option 2. Explore the contents of the snapshot, Metric or Test and find the relevant keys.

To look at all available measurements, you can also:

• Open an existing snapshot file and explore its contents.

• Generate a Report or a Test Suite, include the selected Metric or Test, and get the output as a Python dictionary. You can then explore the keys that contain the metric field names.

Once you identify the specific name of the field you would like to add to a panel, you can pass it as the field_path to the PanelValue parameter.