[ML] Add decision path charts to exploration results table (#73561)

Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>

x-pack/plugins/ml/common/constants/data_frame_analytics.ts

@@ -0,0 +1,7 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+export const DEFAULT_RESULTS_FIELD = 'ml';

x-pack/plugins/ml/common/types/data_frame_analytics.ts

@@ -79,3 +79,9 @@ export interface DataFrameAnalyticsConfig {
   version: string;
   allow_lazy_start?: boolean;
 }
+
+export enum ANALYSIS_CONFIG_TYPE {
+  OUTLIER_DETECTION = 'outlier_detection',
+  REGRESSION = 'regression',
+  CLASSIFICATION = 'classification',
+}

x-pack/plugins/ml/common/types/feature_importance.ts

@@ -0,0 +1,23 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+export interface ClassFeatureImportance {
+  class_name: string | boolean;
+  importance: number;
+}
+export interface FeatureImportance {
+  feature_name: string;
+  importance?: number;
+  classes?: ClassFeatureImportance[];
+}
+
+export interface TopClass {
+  class_name: string;
+  class_probability: number;
+  class_score: number;
+}
+
+export type TopClasses = TopClass[];

x-pack/plugins/ml/common/util/analytics_utils.ts

@@ -0,0 +1,79 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+import {
+  AnalysisConfig,
+  ClassificationAnalysis,
+  OutlierAnalysis,
+  RegressionAnalysis,
+  ANALYSIS_CONFIG_TYPE,
+} from '../types/data_frame_analytics';
+
+export const isOutlierAnalysis = (arg: any): arg is OutlierAnalysis => {
+  const keys = Object.keys(arg);
+  return keys.length === 1 && keys[0] === ANALYSIS_CONFIG_TYPE.OUTLIER_DETECTION;
+};
+
+export const isRegressionAnalysis = (arg: any): arg is RegressionAnalysis => {
+  const keys = Object.keys(arg);
+  return keys.length === 1 && keys[0] === ANALYSIS_CONFIG_TYPE.REGRESSION;
+};
+
+export const isClassificationAnalysis = (arg: any): arg is ClassificationAnalysis => {
+  const keys = Object.keys(arg);
+  return keys.length === 1 && keys[0] === ANALYSIS_CONFIG_TYPE.CLASSIFICATION;
+};
+
+export const getDependentVar = (
+  analysis: AnalysisConfig
+):
+  | RegressionAnalysis['regression']['dependent_variable']
+  | ClassificationAnalysis['classification']['dependent_variable'] => {
+  let depVar = '';
+
+  if (isRegressionAnalysis(analysis)) {
+    depVar = analysis.regression.dependent_variable;
+  }
+
+  if (isClassificationAnalysis(analysis)) {
+    depVar = analysis.classification.dependent_variable;
+  }
+  return depVar;
+};
+
+export const getPredictionFieldName = (
+  analysis: AnalysisConfig
+):
+  | RegressionAnalysis['regression']['prediction_field_name']
+  | ClassificationAnalysis['classification']['prediction_field_name'] => {
+  // If undefined will be defaulted to dependent_variable when config is created
+  let predictionFieldName;
+  if (isRegressionAnalysis(analysis) && analysis.regression.prediction_field_name !== undefined) {
+    predictionFieldName = analysis.regression.prediction_field_name;
+  } else if (
+    isClassificationAnalysis(analysis) &&
+    analysis.classification.prediction_field_name !== undefined
+  ) {
+    predictionFieldName = analysis.classification.prediction_field_name;
+  }
+  return predictionFieldName;
+};
+
+export const getDefaultPredictionFieldName = (analysis: AnalysisConfig) => {
+  return `${getDependentVar(analysis)}_prediction`;
+};
+export const getPredictedFieldName = (
+  resultsField: string,
+  analysis: AnalysisConfig,
+  forSort?: boolean
+) => {
+  // default is 'ml'
+  const predictionFieldName = getPredictionFieldName(analysis);
+  const predictedField = `${resultsField}.${
+    predictionFieldName ? predictionFieldName : getDefaultPredictionFieldName(analysis)
+  }`;
+  return predictedField;
+};

x-pack/plugins/ml/public/application/components/data_grid/common.ts

@@ -119,13 +119,14 @@ export const getDataGridSchemasFromFieldTypes = (fieldTypes: FieldTypes, results
       schema = 'numeric';
     }
 
-    if (
-      field.includes(`${resultsField}.${FEATURE_IMPORTANCE}`) ||
-      field.includes(`${resultsField}.${TOP_CLASSES}`)
-    ) {
+    if (field.includes(`${resultsField}.${TOP_CLASSES}`)) {
       schema = 'json';
     }
 
+    if (field.includes(`${resultsField}.${FEATURE_IMPORTANCE}`)) {
+      schema = 'featureImportance';
+    }
+
     return { id: field, schema, isSortable };
   });
 };
@@ -250,10 +251,6 @@ export const useRenderCellValue = (
         return cellValue ? 'true' : 'false';
       }
 
-      if (typeof cellValue === 'object' && cellValue !== null) {
-        return JSON.stringify(cellValue);
-      }
-
       return cellValue;
     };
   }, [indexPattern?.fields, pagination.pageIndex, pagination.pageSize, tableItems]);

x-pack/plugins/ml/public/application/components/data_grid/data_grid.tsx

@@ -5,8 +5,7 @@
  */
 
 import { isEqual } from 'lodash';
-import React, { memo, useEffect, FC } from 'react';
-
+import React, { memo, useEffect, FC, useMemo } from 'react';
 import { i18n } from '@kbn/i18n';
 
 import {
@@ -24,13 +23,16 @@ import {
 } from '@elastic/eui';
 
 import { CoreSetup } from 'src/core/public';
-
 import { DEFAULT_SAMPLER_SHARD_SIZE } from '../../../../common/constants/field_histograms';
 
-import { INDEX_STATUS } from '../../data_frame_analytics/common';
+import { ANALYSIS_CONFIG_TYPE, INDEX_STATUS } from '../../data_frame_analytics/common';
 
 import { euiDataGridStyle, euiDataGridToolbarSettings } from './common';
 import { UseIndexDataReturnType } from './types';
+import { DecisionPathPopover } from './feature_importance/decision_path_popover';
+import { TopClasses } from '../../../../common/types/feature_importance';
+import { DEFAULT_RESULTS_FIELD } from '../../../../common/constants/data_frame_analytics';
+
 // TODO Fix row hovering + bar highlighting
 // import { hoveredRow$ } from './column_chart';
 
@@ -41,6 +43,9 @@ export const DataGridTitle: FC<{ title: string }> = ({ title }) => (
 );
 
 interface PropsWithoutHeader extends UseIndexDataReturnType {
+  baseline?: number;
+  analysisType?: ANALYSIS_CONFIG_TYPE;
+  resultsField?: string;
   dataTestSubj: string;
   toastNotifications: CoreSetup['notifications']['toasts'];
 }
@@ -60,6 +65,7 @@ type Props = PropsWithHeader | PropsWithoutHeader;
 export const DataGrid: FC<Props> = memo(
   (props) => {
     const {
+      baseline,
       chartsVisible,
       chartsButtonVisible,
       columnsWithCharts,
@@ -80,8 +86,10 @@ export const DataGrid: FC<Props> = memo(
       toastNotifications,
       toggleChartVisibility,
       visibleColumns,
+      predictionFieldName,
+      resultsField,
+      analysisType,
     } = props;
-
     // TODO Fix row hovering + bar highlighting
     // const getRowProps = (item: any) => {
     //   return {
@@ -90,6 +98,45 @@ export const DataGrid: FC<Props> = memo(
     //   };
     // };
 
+    const popOverContent = useMemo(() => {
+      return analysisType === ANALYSIS_CONFIG_TYPE.REGRESSION ||
+        analysisType === ANALYSIS_CONFIG_TYPE.CLASSIFICATION
+        ? {
+            featureImportance: ({ children }: { cellContentsElement: any; children: any }) => {
+              const rowIndex = children?.props?.visibleRowIndex;
+              const row = data[rowIndex];
+              if (!row) return <div />;
+              // if resultsField for some reason is not available then use ml
+              const mlResultsField = resultsField ?? DEFAULT_RESULTS_FIELD;
+              const parsedFIArray = row[mlResultsField].feature_importance;
+              let predictedValue: string | number | undefined;
+              let topClasses: TopClasses = [];
+              if (
+                predictionFieldName !== undefined &&
+                row &&
+                row[mlResultsField][predictionFieldName] !== undefined
+              ) {
+                predictedValue = row[mlResultsField][predictionFieldName];
+                topClasses = row[mlResultsField].top_classes;
+              }
+
+              return (
+                <DecisionPathPopover
+                  analysisType={analysisType}
+                  predictedValue={predictedValue}
+                  baseline={baseline}
+                  featureImportance={parsedFIArray}
+                  topClasses={topClasses}
+                  predictionFieldName={
+                    predictionFieldName ? predictionFieldName.replace('_prediction', '') : undefined
+                  }
+                />
+              );
+            },
+          }
+        : undefined;
+    }, [baseline, data]);
+
     useEffect(() => {
       if (invalidSortingColumnns.length > 0) {
         invalidSortingColumnns.forEach((columnId) => {
@@ -225,6 +272,7 @@ export const DataGrid: FC<Props> = memo(
                   }
                 : {}),
             }}
+            popoverContents={popOverContent}
             pagination={{
               ...pagination,
               pageSizeOptions: [5, 10, 25],

x-pack/plugins/ml/public/application/components/data_grid/feature_importance/decision_path_chart.tsx

@@ -0,0 +1,166 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+import {
+  AnnotationDomainTypes,
+  Axis,
+  AxisStyle,
+  Chart,
+  LineAnnotation,
+  LineAnnotationStyle,
+  LineAnnotationDatum,
+  LineSeries,
+  PartialTheme,
+  Position,
+  RecursivePartial,
+  ScaleType,
+  Settings,
+} from '@elastic/charts';
+import { EuiIcon } from '@elastic/eui';
+
+import React, { useCallback, useMemo } from 'react';
+import { i18n } from '@kbn/i18n';
+import euiVars from '@elastic/eui/dist/eui_theme_light.json';
+import { DecisionPathPlotData } from './use_classification_path_data';
+
+const { euiColorFullShade, euiColorMediumShade } = euiVars;
+const axisColor = euiColorMediumShade;
+
+const baselineStyle: LineAnnotationStyle = {
+  line: {
+    strokeWidth: 1,
+    stroke: euiColorFullShade,
+    opacity: 0.75,
+  },
+  details: {
+    fontFamily: 'Arial',
+    fontSize: 10,
+    fontStyle: 'bold',
+    fill: euiColorMediumShade,
+    padding: 0,
+  },
+};
+
+const axes: RecursivePartial<AxisStyle> = {
+  axisLine: {
+    stroke: axisColor,
+  },
+  tickLabel: {
+    fontSize: 10,
+    fill: axisColor,
+  },
+  tickLine: {
+    stroke: axisColor,
+  },
+  gridLine: {
+    horizontal: {
+      dash: [1, 2],
+    },
+    vertical: {
+      strokeWidth: 0,
+    },
+  },
+};
+const theme: PartialTheme = {
+  axes,
+};
+
+interface DecisionPathChartProps {
+  decisionPathData: DecisionPathPlotData;
+  predictionFieldName?: string;
+  baseline?: number;
+  minDomain: number | undefined;
+  maxDomain: number | undefined;
+}
+
+const DECISION_PATH_MARGIN = 125;
+const DECISION_PATH_ROW_HEIGHT = 10;
+const NUM_PRECISION = 3;
+const AnnotationBaselineMarker = <EuiIcon type="dot" size="m" />;
+
+export const DecisionPathChart = ({
+  decisionPathData,
+  predictionFieldName,
+  minDomain,
+  maxDomain,
+  baseline,
+}: DecisionPathChartProps) => {
+  // adjust the height so it's compact for items with more features
+  const baselineData: LineAnnotationDatum[] = useMemo(
+    () => [
+      {
+        dataValue: baseline,
+        header: baseline ? baseline.toPrecision(NUM_PRECISION) : '',
+        details: i18n.translate(
+          'xpack.ml.dataframe.analytics.explorationResults.decisionPathBaselineText',
+          {
+            defaultMessage:
+              'baseline (average of predictions for all data points in the training data set)',
+          }
+        ),
+      },
+    ],
+    [baseline]
+  );
+  // guarantee up to num_precision significant digits
+  // without having it in scientific notation
+  const tickFormatter = useCallback((d) => Number(d.toPrecision(NUM_PRECISION)).toString(), []);
+
+  return (
+    <Chart
+      size={{ height: DECISION_PATH_MARGIN + decisionPathData.length * DECISION_PATH_ROW_HEIGHT }}
+    >
+      <Settings theme={theme} rotation={90} />
+      {baseline && (
+        <LineAnnotation
+          id="xpack.ml.dataframe.analytics.explorationResults.decisionPathBaseline"
+          domainType={AnnotationDomainTypes.YDomain}
+          dataValues={baselineData}
+          style={baselineStyle}
+          marker={AnnotationBaselineMarker}
+        />
+      )}
+
+      <Axis
+        id={'xpack.ml.dataframe.analytics.explorationResults.decisionPathXAxis'}
+        tickFormat={tickFormatter}
+        title={i18n.translate(
+          'xpack.ml.dataframe.analytics.explorationResults.decisionPathXAxisTitle',
+          {
+            defaultMessage: "Prediction for '{predictionFieldName}'",
+            values: { predictionFieldName },
+          }
+        )}
+        showGridLines={false}
+        position={Position.Top}
+        showOverlappingTicks
+        domain={
+          minDomain && maxDomain
+            ? {
+                min: minDomain,
+                max: maxDomain,
+              }
+            : undefined
+        }
+      />
+      <Axis showGridLines={true} id="left" position={Position.Left} />
+      <LineSeries
+        id={'xpack.ml.dataframe.analytics.explorationResults.decisionPathLine'}
+        name={i18n.translate(
+          'xpack.ml.dataframe.analytics.explorationResults.decisionPathLineTitle',
+          {
+            defaultMessage: 'Prediction',
+          }
+        )}
+        xScaleType={ScaleType.Ordinal}
+        yScaleType={ScaleType.Linear}
+        xAccessor={0}
+        yAccessors={[2]}
+        data={decisionPathData}
+      />
+    </Chart>
+  );
+};

x-pack/plugins/ml/public/application/components/data_grid/feature_importance/decision_path_classification.tsx

@@ -0,0 +1,105 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+import React, { FC, useMemo, useState } from 'react';
+import { i18n } from '@kbn/i18n';
+import { EuiHealth, EuiSpacer, EuiSuperSelect, EuiTitle } from '@elastic/eui';
+import d3 from 'd3';
+import {
+  isDecisionPathData,
+  useDecisionPathData,
+  getStringBasedClassName,
+} from './use_classification_path_data';
+import { FeatureImportance, TopClasses } from '../../../../../common/types/feature_importance';
+import { DecisionPathChart } from './decision_path_chart';
+import { MissingDecisionPathCallout } from './missing_decision_path_callout';
+
+interface ClassificationDecisionPathProps {
+  predictedValue: string | boolean;
+  predictionFieldName?: string;
+  featureImportance: FeatureImportance[];
+  topClasses: TopClasses;
+}
+
+export const ClassificationDecisionPath: FC<ClassificationDecisionPathProps> = ({
+  featureImportance,
+  predictedValue,
+  topClasses,
+  predictionFieldName,
+}) => {
+  const [currentClass, setCurrentClass] = useState<string>(
+    getStringBasedClassName(topClasses[0].class_name)
+  );
+  const { decisionPathData } = useDecisionPathData({
+    featureImportance,
+    predictedValue: currentClass,
+  });
+  const options = useMemo(() => {
+    const predictionValueStr = getStringBasedClassName(predictedValue);
+
+    return Array.isArray(topClasses)
+      ? topClasses.map((c) => {
+          const className = getStringBasedClassName(c.class_name);
+          return {
+            value: className,
+            inputDisplay:
+              className === predictionValueStr ? (
+                <EuiHealth color="success" style={{ lineHeight: 'inherit' }}>
+                  {className}
+                </EuiHealth>
+              ) : (
+                className
+              ),
+          };
+        })
+      : undefined;
+  }, [topClasses, predictedValue]);
+
+  const domain = useMemo(() => {
+    let maxDomain;
+    let minDomain;
+    // if decisionPathData has calculated cumulative path
+    if (decisionPathData && isDecisionPathData(decisionPathData)) {
+      const [min, max] = d3.extent(decisionPathData, (d: [string, number, number]) => d[2]);
+      const buffer = Math.abs(max - min) * 0.1;
+      maxDomain = max + buffer;
+      minDomain = min - buffer;
+    }
+    return { maxDomain, minDomain };
+  }, [decisionPathData]);
+
+  if (!decisionPathData) return <MissingDecisionPathCallout />;
+
+  return (
+    <>
+      <EuiSpacer size={'xs'} />
+      <EuiTitle size={'xxxs'}>
+        <span>
+          {i18n.translate(
+            'xpack.ml.dataframe.analytics.explorationResults.classificationDecisionPathClassNameTitle',
+            {
+              defaultMessage: 'Class name',
+            }
+          )}
+        </span>
+      </EuiTitle>
+      {options !== undefined && (
+        <EuiSuperSelect
+          compressed={true}
+          options={options}
+          valueOfSelected={currentClass}
+          onChange={setCurrentClass}
+        />
+      )}
+      <DecisionPathChart
+        decisionPathData={decisionPathData}
+        predictionFieldName={predictionFieldName}
+        minDomain={domain.minDomain}
+        maxDomain={domain.maxDomain}
+      />
+    </>
+  );
+};

x-pack/plugins/ml/public/application/components/data_grid/feature_importance/decision_path_json_viewer.tsx

@@ -0,0 +1,16 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+import React, { FC } from 'react';
+import { EuiCodeBlock } from '@elastic/eui';
+import { FeatureImportance } from '../../../../../common/types/feature_importance';
+
+interface DecisionPathJSONViewerProps {
+  featureImportance: FeatureImportance[];
+}
+export const DecisionPathJSONViewer: FC<DecisionPathJSONViewerProps> = ({ featureImportance }) => {
+  return <EuiCodeBlock isCopyable={true}>{JSON.stringify(featureImportance)}</EuiCodeBlock>;
+};

x-pack/plugins/ml/public/application/components/data_grid/feature_importance/decision_path_popover.tsx

@@ -0,0 +1,134 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+import React, { FC, useState } from 'react';
+import { EuiLink, EuiTab, EuiTabs, EuiText } from '@elastic/eui';
+import { FormattedMessage } from '@kbn/i18n/react';
+import { RegressionDecisionPath } from './decision_path_regression';
+import { DecisionPathJSONViewer } from './decision_path_json_viewer';
+import { FeatureImportance, TopClasses } from '../../../../../common/types/feature_importance';
+import { ANALYSIS_CONFIG_TYPE } from '../../../data_frame_analytics/common';
+import { ClassificationDecisionPath } from './decision_path_classification';
+import { useMlKibana } from '../../../contexts/kibana';
+
+interface DecisionPathPopoverProps {
+  featureImportance: FeatureImportance[];
+  analysisType: ANALYSIS_CONFIG_TYPE;
+  predictionFieldName?: string;
+  baseline?: number;
+  predictedValue?: number | string | undefined;
+  topClasses?: TopClasses;
+}
+
+enum DECISION_PATH_TABS {
+  CHART = 'decision_path_chart',
+  JSON = 'decision_path_json',
+}
+
+export interface ExtendedFeatureImportance extends FeatureImportance {
+  absImportance?: number;
+}
+
+export const DecisionPathPopover: FC<DecisionPathPopoverProps> = ({
+  baseline,
+  featureImportance,
+  predictedValue,
+  topClasses,
+  analysisType,
+  predictionFieldName,
+}) => {
+  const [selectedTabId, setSelectedTabId] = useState(DECISION_PATH_TABS.CHART);
+  const {
+    services: { docLinks },
+  } = useMlKibana();
+  const { ELASTIC_WEBSITE_URL, DOC_LINK_VERSION } = docLinks;
+
+  if (featureImportance.length < 2) {
+    return <DecisionPathJSONViewer featureImportance={featureImportance} />;
+  }
+
+  const tabs = [
+    {
+      id: DECISION_PATH_TABS.CHART,
+      name: (
+        <FormattedMessage
+          id="xpack.ml.dataframe.analytics.explorationResults.decisionPathPlotTab"
+          defaultMessage="Decision plot"
+        />
+      ),
+    },
+    {
+      id: DECISION_PATH_TABS.JSON,
+      name: (
+        <FormattedMessage
+          id="xpack.ml.dataframe.analytics.explorationResults.decisionPathJSONTab"
+          defaultMessage="JSON"
+        />
+      ),
+    },
+  ];
+
+  return (
+    <>
+      <div style={{ display: 'flex', width: 300 }}>
+        <EuiTabs size={'s'}>
+          {tabs.map((tab) => (
+            <EuiTab
+              isSelected={tab.id === selectedTabId}
+              onClick={() => setSelectedTabId(tab.id)}
+              key={tab.id}
+            >
+              {tab.name}
+            </EuiTab>
+          ))}
+        </EuiTabs>
+      </div>
+      {selectedTabId === DECISION_PATH_TABS.CHART && (
+        <>
+          <EuiText size={'xs'} color="subdued" style={{ paddingTop: 5 }}>
+            <FormattedMessage
+              id="xpack.ml.dataframe.analytics.explorationResults.decisionPathPlotHelpText"
+              defaultMessage="SHAP decision plots use {linkedFeatureImportanceValues} to show how models arrive at the predicted value for '{predictionFieldName}'."
+              values={{
+                predictionFieldName,
+                linkedFeatureImportanceValues: (
+                  <EuiLink
+                    href={`${ELASTIC_WEBSITE_URL}guide/en/machine-learning/${DOC_LINK_VERSION}/ml-feature-importance.html`}
+                    target="_blank"
+                  >
+                    <FormattedMessage
+                      id="xpack.ml.dataframe.analytics.explorationResults.linkedFeatureImportanceValues"
+                      defaultMessage="feature importance values"
+                    />
+                  </EuiLink>
+                ),
+              }}
+            />
+          </EuiText>
+          {analysisType === ANALYSIS_CONFIG_TYPE.CLASSIFICATION && (
+            <ClassificationDecisionPath
+              featureImportance={featureImportance}
+              topClasses={topClasses as TopClasses}
+              predictedValue={predictedValue as string}
+              predictionFieldName={predictionFieldName}
+            />
+          )}
+          {analysisType === ANALYSIS_CONFIG_TYPE.REGRESSION && (
+            <RegressionDecisionPath
+              featureImportance={featureImportance}
+              baseline={baseline}
+              predictedValue={predictedValue as number}
+              predictionFieldName={predictionFieldName}
+            />
+          )}
+        </>
+      )}
+      {selectedTabId === DECISION_PATH_TABS.JSON && (
+        <DecisionPathJSONViewer featureImportance={featureImportance} />
+      )}
+    </>
+  );
+};

x-pack/plugins/ml/public/application/components/data_grid/feature_importance/decision_path_regression.tsx

@@ -0,0 +1,79 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+import React, { FC, useMemo } from 'react';
+import { EuiCallOut } from '@elastic/eui';
+import { FormattedMessage } from '@kbn/i18n/react';
+import d3 from 'd3';
+import { FeatureImportance, TopClasses } from '../../../../../common/types/feature_importance';
+import { useDecisionPathData, isDecisionPathData } from './use_classification_path_data';
+import { DecisionPathChart } from './decision_path_chart';
+import { MissingDecisionPathCallout } from './missing_decision_path_callout';
+
+interface RegressionDecisionPathProps {
+  predictionFieldName?: string;
+  baseline?: number;
+  predictedValue?: number | undefined;
+  featureImportance: FeatureImportance[];
+  topClasses?: TopClasses;
+}
+
+export const RegressionDecisionPath: FC<RegressionDecisionPathProps> = ({
+  baseline,
+  featureImportance,
+  predictedValue,
+  predictionFieldName,
+}) => {
+  const { decisionPathData } = useDecisionPathData({
+    baseline,
+    featureImportance,
+    predictedValue,
+  });
+  const domain = useMemo(() => {
+    let maxDomain;
+    let minDomain;
+    // if decisionPathData has calculated cumulative path
+    if (decisionPathData && isDecisionPathData(decisionPathData)) {
+      const [min, max] = d3.extent(decisionPathData, (d: [string, number, number]) => d[2]);
+      maxDomain = max;
+      minDomain = min;
+      const buffer = Math.abs(maxDomain - minDomain) * 0.1;
+      maxDomain =
+        (typeof baseline === 'number' ? Math.max(maxDomain, baseline) : maxDomain) + buffer;
+      minDomain =
+        (typeof baseline === 'number' ? Math.min(minDomain, baseline) : minDomain) - buffer;
+    }
+    return { maxDomain, minDomain };
+  }, [decisionPathData, baseline]);
+
+  if (!decisionPathData) return <MissingDecisionPathCallout />;
+
+  return (
+    <>
+      {baseline === undefined && (
+        <EuiCallOut
+          size={'s'}
+          heading={'p'}
+          title={
+            <FormattedMessage
+              id="xpack.ml.dataframe.analytics.explorationResults.missingBaselineCallout"
+              defaultMessage="Unable to calculate baseline value, which might result in a shifted decision path."
+            />
+          }
+          color="warning"
+          iconType="alert"
+        />
+      )}
+      <DecisionPathChart
+        decisionPathData={decisionPathData}
+        predictionFieldName={predictionFieldName}
+        minDomain={domain.minDomain}
+        maxDomain={domain.maxDomain}
+        baseline={baseline}
+      />
+    </>
+  );
+};

x-pack/plugins/ml/public/application/components/data_grid/feature_importance/missing_decision_path_callout.tsx

@@ -0,0 +1,20 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+import React from 'react';
+import { EuiCallOut } from '@elastic/eui';
+import { FormattedMessage } from '@kbn/i18n/react';
+
+export const MissingDecisionPathCallout = () => {
+  return (
+    <EuiCallOut color={'warning'}>
+      <FormattedMessage
+        id="xpack.ml.dataframe.analytics.explorationResults.regressionDecisionPathDataMissingCallout"
+        defaultMessage="No decision path data available."
+      />
+    </EuiCallOut>
+  );
+};

x-pack/plugins/ml/public/application/components/data_grid/feature_importance/use_classification_path_data.tsx

@@ -0,0 +1,173 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+import { useMemo } from 'react';
+import { i18n } from '@kbn/i18n';
+import { FeatureImportance, TopClasses } from '../../../../../common/types/feature_importance';
+import { ExtendedFeatureImportance } from './decision_path_popover';
+
+export type DecisionPathPlotData = Array<[string, number, number]>;
+
+interface UseDecisionPathDataParams {
+  featureImportance: FeatureImportance[];
+  baseline?: number;
+  predictedValue?: string | number | undefined;
+  topClasses?: TopClasses;
+}
+
+interface RegressionDecisionPathProps {
+  baseline?: number;
+  predictedValue?: number | undefined;
+  featureImportance: FeatureImportance[];
+  topClasses?: TopClasses;
+}
+const FEATURE_NAME = 'feature_name';
+const FEATURE_IMPORTANCE = 'importance';
+
+export const isDecisionPathData = (decisionPathData: any): boolean => {
+  return (
+    Array.isArray(decisionPathData) &&
+    decisionPathData.length > 0 &&
+    decisionPathData[0].length === 3
+  );
+};
+
+// cast to 'True' | 'False' | value to match Eui display
+export const getStringBasedClassName = (v: string | boolean | undefined | number): string => {
+  if (v === undefined) {
+    return '';
+  }
+  if (typeof v === 'boolean') {
+    return v ? 'True' : 'False';
+  }
+  if (typeof v === 'number') {
+    return v.toString();
+  }
+  return v;
+};
+
+export const useDecisionPathData = ({
+  baseline,
+  featureImportance,
+  predictedValue,
+}: UseDecisionPathDataParams): { decisionPathData: DecisionPathPlotData | undefined } => {
+  const decisionPathData = useMemo(() => {
+    return baseline
+      ? buildRegressionDecisionPathData({
+          baseline,
+          featureImportance,
+          predictedValue: predictedValue as number | undefined,
+        })
+      : buildClassificationDecisionPathData({
+          featureImportance,
+          currentClass: predictedValue as string | undefined,
+        });
+  }, [baseline, featureImportance, predictedValue]);
+
+  return { decisionPathData };
+};
+
+export const buildDecisionPathData = (featureImportance: ExtendedFeatureImportance[]) => {
+  const finalResult: DecisionPathPlotData = featureImportance
+    // sort so absolute importance so it goes from bottom (baseline) to top
+    .sort(
+      (a: ExtendedFeatureImportance, b: ExtendedFeatureImportance) =>
+        b.absImportance! - a.absImportance!
+    )
+    .map((d) => [d[FEATURE_NAME] as string, d[FEATURE_IMPORTANCE] as number, NaN]);
+
+  // start at the baseline and end at predicted value
+  // for regression, cumulativeSum should add up to baseline
+  let cumulativeSum = 0;
+  for (let i = featureImportance.length - 1; i >= 0; i--) {
+    cumulativeSum += finalResult[i][1];
+    finalResult[i][2] = cumulativeSum;
+  }
+  return finalResult;
+};
+export const buildRegressionDecisionPathData = ({
+  baseline,
+  featureImportance,
+  predictedValue,
+}: RegressionDecisionPathProps): DecisionPathPlotData | undefined => {
+  let mappedFeatureImportance: ExtendedFeatureImportance[] = featureImportance;
+  mappedFeatureImportance = mappedFeatureImportance.map((d) => ({
+    ...d,
+    absImportance: Math.abs(d[FEATURE_IMPORTANCE] as number),
+  }));
+
+  if (baseline && predictedValue !== undefined && Number.isFinite(predictedValue)) {
+    // get the adjusted importance needed for when # of fields included in c++ analysis != max allowed
+    // if num fields included = num features allowed exactly, adjustedImportance should be 0
+    const adjustedImportance =
+      predictedValue -
+      mappedFeatureImportance.reduce(
+        (accumulator, currentValue) => accumulator + currentValue.importance!,
+        0
+      ) -
+      baseline;
+
+    mappedFeatureImportance.push({
+      [FEATURE_NAME]: i18n.translate(
+        'xpack.ml.dataframe.analytics.decisionPathFeatureBaselineTitle',
+        {
+          defaultMessage: 'baseline',
+        }
+      ),
+      [FEATURE_IMPORTANCE]: baseline,
+      absImportance: -1,
+    });
+
+    // if the difference is small enough then no need to plot the residual feature importance
+    if (Math.abs(adjustedImportance) > 1e-5) {
+      mappedFeatureImportance.push({
+        [FEATURE_NAME]: i18n.translate(
+          'xpack.ml.dataframe.analytics.decisionPathFeatureOtherTitle',
+          {
+            defaultMessage: 'other',
+          }
+        ),
+        [FEATURE_IMPORTANCE]: adjustedImportance,
+        absImportance: 0, // arbitrary importance so this will be of higher importance than baseline
+      });
+    }
+  }
+  const filteredFeatureImportance = mappedFeatureImportance.filter(
+    (f) => f !== undefined
+  ) as ExtendedFeatureImportance[];
+
+  return buildDecisionPathData(filteredFeatureImportance);
+};
+
+export const buildClassificationDecisionPathData = ({
+  featureImportance,
+  currentClass,
+}: {
+  featureImportance: FeatureImportance[];
+  currentClass: string | undefined;
+}): DecisionPathPlotData | undefined => {
+  if (currentClass === undefined) return [];
+  const mappedFeatureImportance: Array<
+    ExtendedFeatureImportance | undefined
+  > = featureImportance.map((feature) => {
+    const classFeatureImportance = Array.isArray(feature.classes)
+      ? feature.classes.find((c) => getStringBasedClassName(c.class_name) === currentClass)
+      : feature;
+    if (classFeatureImportance && typeof classFeatureImportance[FEATURE_IMPORTANCE] === 'number') {
+      return {
+        [FEATURE_NAME]: feature[FEATURE_NAME],
+        [FEATURE_IMPORTANCE]: classFeatureImportance[FEATURE_IMPORTANCE],
+        absImportance: Math.abs(classFeatureImportance[FEATURE_IMPORTANCE] as number),
+      };
+    }
+    return undefined;
+  });
+  const filteredFeatureImportance = mappedFeatureImportance.filter(
+    (f) => f !== undefined
+  ) as ExtendedFeatureImportance[];
+
+  return buildDecisionPathData(filteredFeatureImportance);
+};

x-pack/plugins/ml/public/application/components/data_grid/types.ts

@@ -74,6 +74,9 @@ export interface UseIndexDataReturnType
     | 'tableItems'
     | 'toggleChartVisibility'
     | 'visibleColumns'
+    | 'baseline'
+    | 'predictionFieldName'
+    | 'resultsField'
   > {
   renderCellValue: RenderCellValue;
 }
@@ -105,4 +108,7 @@ export interface UseDataGridReturnType {
   tableItems: DataGridItem[];
   toggleChartVisibility: () => void;
   visibleColumns: ColumnId[];
+  baseline?: number;
+  predictionFieldName?: string;
+  resultsField?: string;
 }

x-pack/plugins/ml/public/application/data_frame_analytics/common/analytics.ts

@@ -15,18 +15,19 @@ import { SavedSearchQuery } from '../../contexts/ml';
 import {
   AnalysisConfig,
   ClassificationAnalysis,
-  OutlierAnalysis,
   RegressionAnalysis,
+  ANALYSIS_CONFIG_TYPE,
 } from '../../../../common/types/data_frame_analytics';
-
+import {
+  isOutlierAnalysis,
+  isRegressionAnalysis,
+  isClassificationAnalysis,
+  getPredictionFieldName,
+  getDependentVar,
+  getPredictedFieldName,
+} from '../../../../common/util/analytics_utils';
 export type IndexPattern = string;
 
-export enum ANALYSIS_CONFIG_TYPE {
-  OUTLIER_DETECTION = 'outlier_detection',
-  REGRESSION = 'regression',
-  CLASSIFICATION = 'classification',
-}
-
 export enum ANALYSIS_ADVANCED_FIELDS {
   ETA = 'eta',
   FEATURE_BAG_FRACTION = 'feature_bag_fraction',
@@ -156,23 +157,6 @@ export const getAnalysisType = (analysis: AnalysisConfig): string => {
   return 'unknown';
 };
 
-export const getDependentVar = (
-  analysis: AnalysisConfig
-):
-  | RegressionAnalysis['regression']['dependent_variable']
-  | ClassificationAnalysis['classification']['dependent_variable'] => {
-  let depVar = '';
-
-  if (isRegressionAnalysis(analysis)) {
-    depVar = analysis.regression.dependent_variable;
-  }
-
-  if (isClassificationAnalysis(analysis)) {
-    depVar = analysis.classification.dependent_variable;
-  }
-  return depVar;
-};
-
 export const getTrainingPercent = (
   analysis: AnalysisConfig
 ):
@@ -190,24 +174,6 @@ export const getTrainingPercent = (
   return trainingPercent;
 };
 
-export const getPredictionFieldName = (
-  analysis: AnalysisConfig
-):
-  | RegressionAnalysis['regression']['prediction_field_name']
-  | ClassificationAnalysis['classification']['prediction_field_name'] => {
-  // If undefined will be defaulted to dependent_variable when config is created
-  let predictionFieldName;
-  if (isRegressionAnalysis(analysis) && analysis.regression.prediction_field_name !== undefined) {
-    predictionFieldName = analysis.regression.prediction_field_name;
-  } else if (
-    isClassificationAnalysis(analysis) &&
-    analysis.classification.prediction_field_name !== undefined
-  ) {
-    predictionFieldName = analysis.classification.prediction_field_name;
-  }
-  return predictionFieldName;
-};
-
 export const getNumTopClasses = (
   analysis: AnalysisConfig
 ): ClassificationAnalysis['classification']['num_top_classes'] => {
@@ -238,35 +204,6 @@ export const getNumTopFeatureImportanceValues = (
   return numTopFeatureImportanceValues;
 };
 
-export const getPredictedFieldName = (
-  resultsField: string,
-  analysis: AnalysisConfig,
-  forSort?: boolean
-) => {
-  // default is 'ml'
-  const predictionFieldName = getPredictionFieldName(analysis);
-  const defaultPredictionField = `${getDependentVar(analysis)}_prediction`;
-  const predictedField = `${resultsField}.${
-    predictionFieldName ? predictionFieldName : defaultPredictionField
-  }`;
-  return predictedField;
-};
-
-export const isOutlierAnalysis = (arg: any): arg is OutlierAnalysis => {
-  const keys = Object.keys(arg);
-  return keys.length === 1 && keys[0] === ANALYSIS_CONFIG_TYPE.OUTLIER_DETECTION;
-};
-
-export const isRegressionAnalysis = (arg: any): arg is RegressionAnalysis => {
-  const keys = Object.keys(arg);
-  return keys.length === 1 && keys[0] === ANALYSIS_CONFIG_TYPE.REGRESSION;
-};
-
-export const isClassificationAnalysis = (arg: any): arg is ClassificationAnalysis => {
-  const keys = Object.keys(arg);
-  return keys.length === 1 && keys[0] === ANALYSIS_CONFIG_TYPE.CLASSIFICATION;
-};
-
 export const isResultsSearchBoolQuery = (arg: any): arg is ResultsSearchBoolQuery => {
   if (arg === undefined) return false;
   const keys = Object.keys(arg);
@@ -607,3 +544,13 @@ export const loadDocsCount = async ({
     };
   }
 };
+
+export {
+  isOutlierAnalysis,
+  isRegressionAnalysis,
+  isClassificationAnalysis,
+  getPredictionFieldName,
+  ANALYSIS_CONFIG_TYPE,
+  getDependentVar,
+  getPredictedFieldName,
+};

x-pack/plugins/ml/public/application/data_frame_analytics/common/constants.ts

@@ -3,8 +3,6 @@
  * or more contributor license agreements. Licensed under the Elastic License;
  * you may not use this file except in compliance with the Elastic License.
  */
-
-export const DEFAULT_RESULTS_FIELD = 'ml';
 export const FEATURE_IMPORTANCE = 'feature_importance';
 export const FEATURE_INFLUENCE = 'feature_influence';
 export const TOP_CLASSES = 'top_classes';

x-pack/plugins/ml/public/application/data_frame_analytics/common/fields.ts

@@ -4,17 +4,16 @@
  * you may not use this file except in compliance with the Elastic License.
  */
 
+import { getNumTopClasses, getNumTopFeatureImportanceValues } from './analytics';
+import { Field } from '../../../../common/types/fields';
 import {
-  getNumTopClasses,
-  getNumTopFeatureImportanceValues,
   getPredictedFieldName,
   getDependentVar,
   getPredictionFieldName,
   isClassificationAnalysis,
   isOutlierAnalysis,
   isRegressionAnalysis,
-} from './analytics';
-import { Field } from '../../../../common/types/fields';
+} from '../../../../common/util/analytics_utils';
 import { ES_FIELD_TYPES, KBN_FIELD_TYPES } from '../../../../../../../src/plugins/data/public';
 import { newJobCapsService } from '../../services/new_job_capabilities_service';
 

x-pack/plugins/ml/public/application/data_frame_analytics/pages/analytics_exploration/components/classification_exploration/classification_exploration.tsx

@@ -9,7 +9,6 @@ import React, { FC } from 'react';
 import { i18n } from '@kbn/i18n';
 
 import { ExplorationPageWrapper } from '../exploration_page_wrapper';
-
 import { EvaluatePanel } from './evaluate_panel';
 
 interface Props {

x-pack/plugins/ml/public/application/data_frame_analytics/pages/analytics_exploration/components/exploration_page_wrapper/exploration_page_wrapper.tsx

@@ -51,7 +51,6 @@ export const ExplorationPageWrapper: FC<Props> = ({ jobId, title, EvaluatePanel
       />
     );
   }
-
   return (
     <>
       {isLoadingJobConfig === true && jobConfig === undefined && <LoadingPanel />}

x-pack/plugins/ml/public/application/data_frame_analytics/pages/analytics_exploration/components/exploration_results_table/exploration_results_table.tsx

@@ -28,6 +28,8 @@ import {
   INDEX_STATUS,
   SEARCH_SIZE,
   defaultSearchQuery,
+  getAnalysisType,
+  ANALYSIS_CONFIG_TYPE,
 } from '../../../../common';
 import { getTaskStateBadge } from '../../../analytics_management/components/analytics_list/use_columns';
 import { DATA_FRAME_TASK_STATE } from '../../../analytics_management/components/analytics_list/common';
@@ -36,6 +38,7 @@ import { ExplorationQueryBar } from '../exploration_query_bar';
 import { IndexPatternPrompt } from '../index_pattern_prompt';
 
 import { useExplorationResults } from './use_exploration_results';
+import { useMlKibana } from '../../../../../contexts/kibana';
 
 const showingDocs = i18n.translate(
   'xpack.ml.dataframe.analytics.explorationResults.documentsShownHelpText',
@@ -70,18 +73,27 @@ export const ExplorationResultsTable: FC<Props> = React.memo(
     setEvaluateSearchQuery,
     title,
   }) => {
+    const {
+      services: {
+        mlServices: { mlApiServices },
+      },
+    } = useMlKibana();
     const [searchQuery, setSearchQuery] = useState<SavedSearchQuery>(defaultSearchQuery);
 
     useEffect(() => {
       setEvaluateSearchQuery(searchQuery);
     }, [JSON.stringify(searchQuery)]);
 
+    const analysisType = getAnalysisType(jobConfig.analysis);
+
     const classificationData = useExplorationResults(
       indexPattern,
       jobConfig,
       searchQuery,
-      getToastNotifications()
+      getToastNotifications(),
+      mlApiServices
     );
+
     const docFieldsCount = classificationData.columnsWithCharts.length;
     const {
       columnsWithCharts,
@@ -94,7 +106,6 @@ export const ExplorationResultsTable: FC<Props> = React.memo(
     if (jobConfig === undefined || classificationData === undefined) {
       return null;
     }
-
     // if it's a searchBar syntax error leave the table visible so they can try again
     if (status === INDEX_STATUS.ERROR && !errorMessage.includes('failed to create query')) {
       return (
@@ -184,6 +195,7 @@ export const ExplorationResultsTable: FC<Props> = React.memo(
                 {...classificationData}
                 dataTestSubj="mlExplorationDataGrid"
                 toastNotifications={getToastNotifications()}
+                analysisType={(analysisType as unknown) as ANALYSIS_CONFIG_TYPE}
               />
             </EuiFlexItem>
           </EuiFlexGroup>

x-pack/plugins/ml/public/application/data_frame_analytics/pages/analytics_exploration/components/exploration_results_table/use_exploration_results.ts

@@ -4,12 +4,14 @@
  * you may not use this file except in compliance with the Elastic License.
  */
 
-import { useEffect, useMemo } from 'react';
+import { useCallback, useEffect, useMemo, useState } from 'react';
 
 import { EuiDataGridColumn } from '@elastic/eui';
 
 import { CoreSetup } from 'src/core/public';
 
+import { i18n } from '@kbn/i18n';
+import { MlApiServices } from '../../../../../services/ml_api_service';
 import { IndexPattern } from '../../../../../../../../../../src/plugins/data/public';
 
 import { DataLoader } from '../../../../../datavisualizer/index_based/data_loader';
@@ -23,21 +25,26 @@ import {
   UseIndexDataReturnType,
 } from '../../../../../components/data_grid';
 import { SavedSearchQuery } from '../../../../../contexts/ml';
-
 import { getIndexData, getIndexFields, DataFrameAnalyticsConfig } from '../../../../common';
 import {
-  DEFAULT_RESULTS_FIELD,
-  FEATURE_IMPORTANCE,
-  TOP_CLASSES,
-} from '../../../../common/constants';
+  getPredictionFieldName,
+  getDefaultPredictionFieldName,
+} from '../../../../../../../common/util/analytics_utils';
+import { FEATURE_IMPORTANCE, TOP_CLASSES } from '../../../../common/constants';
+import { DEFAULT_RESULTS_FIELD } from '../../../../../../../common/constants/data_frame_analytics';
 import { sortExplorationResultsFields, ML__ID_COPY } from '../../../../common/fields';
+import { isRegressionAnalysis } from '../../../../common/analytics';
+import { extractErrorMessage } from '../../../../../../../common/util/errors';
 
 export const useExplorationResults = (
   indexPattern: IndexPattern | undefined,
   jobConfig: DataFrameAnalyticsConfig | undefined,
   searchQuery: SavedSearchQuery,
-  toastNotifications: CoreSetup['notifications']['toasts']
+  toastNotifications: CoreSetup['notifications']['toasts'],
+  mlApiServices: MlApiServices
 ): UseIndexDataReturnType => {
+  const [baseline, setBaseLine] = useState();
+
   const needsDestIndexFields =
     indexPattern !== undefined && indexPattern.title === jobConfig?.source.index[0];
 
@@ -52,7 +59,6 @@ export const useExplorationResults = (
       )
     );
   }
-
   const dataGrid = useDataGrid(
     columns,
     25,
@@ -107,16 +113,60 @@ export const useExplorationResults = (
     jobConfig?.dest.index,
     JSON.stringify([searchQuery, dataGrid.visibleColumns]),
   ]);
+  const predictionFieldName = useMemo(() => {
+    if (jobConfig) {
+      return (
+        getPredictionFieldName(jobConfig.analysis) ??
+        getDefaultPredictionFieldName(jobConfig.analysis)
+      );
+    }
+    return undefined;
+  }, [jobConfig]);
 
+  const getAnalyticsBaseline = useCallback(async () => {
+    try {
+      if (
+        jobConfig !== undefined &&
+        jobConfig.analysis !== undefined &&
+        isRegressionAnalysis(jobConfig.analysis)
+      ) {
+        const result = await mlApiServices.dataFrameAnalytics.getAnalyticsBaseline(jobConfig.id);
+        if (result?.baseline) {
+          setBaseLine(result.baseline);
+        }
+      }
+    } catch (e) {
+      const error = extractErrorMessage(e);
+
+      toastNotifications.addDanger({
+        title: i18n.translate(
+          'xpack.ml.dataframe.analytics.explorationResults.baselineErrorMessageToast',
+          {
+            defaultMessage: 'An error occurred getting feature importance baseline',
+          }
+        ),
+        text: error,
+      });
+    }
+  }, [mlApiServices, jobConfig]);
+
+  useEffect(() => {
+    getAnalyticsBaseline();
+  }, [jobConfig]);
+
+  const resultsField = jobConfig?.dest.results_field ?? DEFAULT_RESULTS_FIELD;
   const renderCellValue = useRenderCellValue(
     indexPattern,
     dataGrid.pagination,
     dataGrid.tableItems,
-    jobConfig?.dest.results_field ?? DEFAULT_RESULTS_FIELD
+    resultsField
   );
 
   return {
     ...dataGrid,
     renderCellValue,
+    baseline,
+    predictionFieldName,
+    resultsField,
   };
 };

x-pack/plugins/ml/public/application/data_frame_analytics/pages/analytics_exploration/components/outlier_exploration/use_outlier_data.ts

@@ -29,7 +29,8 @@ import { SavedSearchQuery } from '../../../../../contexts/ml';
 import { getToastNotifications } from '../../../../../util/dependency_cache';
 
 import { getIndexData, getIndexFields, DataFrameAnalyticsConfig } from '../../../../common';
-import { DEFAULT_RESULTS_FIELD, FEATURE_INFLUENCE } from '../../../../common/constants';
+import { FEATURE_INFLUENCE } from '../../../../common/constants';
+import { DEFAULT_RESULTS_FIELD } from '../../../../../../../common/constants/data_frame_analytics';
 import { sortExplorationResultsFields, ML__ID_COPY } from '../../../../common/fields';
 
 import { getFeatureCount, getOutlierScoreFieldName } from './common';

x-pack/plugins/ml/public/application/data_frame_analytics/pages/analytics_management/components/action_clone/clone_action_name.tsx

@@ -12,7 +12,7 @@ import { IIndexPattern } from 'src/plugins/data/common';
 import { DeepReadonly } from '../../../../../../../common/types/common';
 import { DataFrameAnalyticsConfig, isOutlierAnalysis } from '../../../../common';
 import { isClassificationAnalysis, isRegressionAnalysis } from '../../../../common/analytics';
-import { DEFAULT_RESULTS_FIELD } from '../../../../common/constants';
+import { DEFAULT_RESULTS_FIELD } from '../../../../../../../common/constants/data_frame_analytics';
 import { useMlKibana, useNavigateToPath } from '../../../../../contexts/kibana';
 import { DEFAULT_NUM_TOP_FEATURE_IMPORTANCE_VALUES } from '../../hooks/use_create_analytics_form';
 import { State } from '../../hooks/use_create_analytics_form/state';

x-pack/plugins/ml/public/application/services/ml_api_service/data_frame_analytics.ts

@@ -135,4 +135,10 @@ export const dataFrameAnalytics = {
       method: 'GET',
     });
   },
+  getAnalyticsBaseline(analyticsId: string) {
+    return http<any>({
+      path: `${basePath()}/data_frame/analytics/${analyticsId}/baseline`,
+      method: 'POST',
+    });
+  },
 };

x-pack/plugins/ml/server/models/data_frame_analytics/feature_importance.ts

@@ -0,0 +1,69 @@
+/*
+ * Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
+ * or more contributor license agreements. Licensed under the Elastic License;
+ * you may not use this file except in compliance with the Elastic License.
+ */
+
+import { IScopedClusterClient } from 'kibana/server';
+import {
+  getDefaultPredictionFieldName,
+  getPredictionFieldName,
+  isRegressionAnalysis,
+} from '../../../common/util/analytics_utils';
+import { DEFAULT_RESULTS_FIELD } from '../../../common/constants/data_frame_analytics';
+// Obtains data for the data frame analytics feature importance functionalities
+// such as baseline, decision paths, or importance summary.
+export function analyticsFeatureImportanceProvider({
+  asInternalUser,
+  asCurrentUser,
+}: IScopedClusterClient) {
+  async function getRegressionAnalyticsBaseline(analyticsId: string): Promise<number | undefined> {
+    const { body } = await asInternalUser.ml.getDataFrameAnalytics({
+      id: analyticsId,
+    });
+    const jobConfig = body.data_frame_analytics[0];
+    if (!isRegressionAnalysis) return undefined;
+    const destinationIndex = jobConfig.dest.index;
+    const predictionFieldName = getPredictionFieldName(jobConfig.analysis);
+    const mlResultsField = jobConfig.dest?.results_field ?? DEFAULT_RESULTS_FIELD;
+    const predictedField = `${mlResultsField}.${
+      predictionFieldName ? predictionFieldName : getDefaultPredictionFieldName(jobConfig.analysis)
+    }`;
+    const isTrainingField = `${mlResultsField}.is_training`;
+
+    const params = {
+      index: destinationIndex,
+      size: 0,
+      body: {
+        query: {
+          bool: {
+            filter: [
+              {
+                term: {
+                  [isTrainingField]: true,
+                },
+              },
+            ],
+          },
+        },
+        aggs: {
+          featureImportanceBaseline: {
+            avg: {
+              field: predictedField,
+            },
+          },
+        },
+      },
+    };
+    let baseline;
+    const { body: aggregationResult } = await asCurrentUser.search(params);
+    if (aggregationResult) {
+      baseline = aggregationResult.aggregations.featureImportanceBaseline.value;
+    }
+    return baseline;
+  }
+
+  return {
+    getRegressionAnalyticsBaseline,
+  };
+}

x-pack/plugins/ml/server/routes/data_frame_analytics.ts

@@ -20,6 +20,7 @@ import {
 import { IndexPatternHandler } from '../models/data_frame_analytics/index_patterns';
 import { DeleteDataFrameAnalyticsWithIndexStatus } from '../../common/types/data_frame_analytics';
 import { getAuthorizationHeader } from '../lib/request_authorization';
+import { analyticsFeatureImportanceProvider } from '../models/data_frame_analytics/feature_importance';
 
 function getIndexPatternId(context: RequestHandlerContext, patternName: string) {
   const iph = new IndexPatternHandler(context.core.savedObjects.client);
@@ -545,4 +546,38 @@ export function dataFrameAnalyticsRoutes({ router, mlLicense }: RouteInitializat
       }
     })
   );
+
+  /**
+   * @apiGroup DataFrameAnalytics
+   *
+   * @api {get} /api/ml/data_frame/analytics/baseline Get analytics's feature importance baseline
+   * @apiName GetDataFrameAnalyticsBaseline
+   * @apiDescription Returns the baseline for data frame analytics job.
+   *
+   * @apiSchema (params) analyticsIdSchema
+   */
+  router.post(
+    {
+      path: '/api/ml/data_frame/analytics/{analyticsId}/baseline',
+      validate: {
+        params: analyticsIdSchema,
+      },
+      options: {
+        tags: ['access:ml:canGetDataFrameAnalytics'],
+      },
+    },
+    mlLicense.fullLicenseAPIGuard(async ({ client, request, response }) => {
+      try {
+        const { analyticsId } = request.params;
+        const { getRegressionAnalyticsBaseline } = analyticsFeatureImportanceProvider(client);
+        const baseline = await getRegressionAnalyticsBaseline(analyticsId);
+
+        return response.ok({
+          body: { baseline },
+        });
+      } catch (e) {
+        return response.customError(wrapError(e));
+      }
+    })
+  );
 }

x-pack/test/functional/services/ml/data_frame_analytics_creation.ts

@@ -10,25 +10,9 @@ import { FtrProviderContext } from '../../ftr_provider_context';
 import { MlCommonUI } from './common_ui';
 import { MlApi } from './api';
 import {
-  ClassificationAnalysis,
-  RegressionAnalysis,
-} from '../../../../plugins/ml/common/types/data_frame_analytics';
-
-enum ANALYSIS_CONFIG_TYPE {
-  OUTLIER_DETECTION = 'outlier_detection',
-  REGRESSION = 'regression',
-  CLASSIFICATION = 'classification',
-}
-
-const isRegressionAnalysis = (arg: any): arg is RegressionAnalysis => {
-  const keys = Object.keys(arg);
-  return keys.length === 1 && keys[0] === ANALYSIS_CONFIG_TYPE.REGRESSION;
-};
-
-const isClassificationAnalysis = (arg: any): arg is ClassificationAnalysis => {
-  const keys = Object.keys(arg);
-  return keys.length === 1 && keys[0] === ANALYSIS_CONFIG_TYPE.CLASSIFICATION;
-};
+  isRegressionAnalysis,
+  isClassificationAnalysis,
+} from '../../../../plugins/ml/common/util/analytics_utils';
 
 export function MachineLearningDataFrameAnalyticsCreationProvider(
   { getService }: FtrProviderContext,