tinybase 0.0.0 → 0.9.0

package/lib/metrics.d.ts

@@ -0,0 +1,728 @@
+/**
+ * The metrics module of the TinyBase project provides the ability to create
+ * and track metrics and aggregates of the data in Store objects.
+ *
+ * The main entry point to this module is the createMetrics function, which
+ * returns a new Metrics object. From there, you can create new metric
+ * definitions, access the values of those metrics directly, and register
+ * listeners for when they change.
+ *
+ * @packageDocumentation
+ * @module metrics
+ */
+
+import {GetCell, Store} from './store.d';
+import {Id, IdOrNull, Ids} from './common.d';
+
+/**
+ * The Metric type is simply an alias, but represents a number formed by
+ * aggregating multiple other numbers together.
+ *
+ * @category Metric
+ */
+export type Metric = number;
+
+/**
+ * The Aggregate type describes a custom function that takes an array or numbers
+ * and returns an aggregate that is used as a Metric.
+ *
+ * There are a number of common predefined aggregators, such as for counting,
+ * summing, and averaging values. This type is instead used for when you wish to
+ * use a more complex aggregation of your own devising. See the
+ * setMetricDefinition method for more examples.
+ *
+ * @category Aggregators
+ */
+export type Aggregate = (numbers: number[], length: number) => Metric;
+
+/**
+ * The AggregateAdd type describes a function that can be used to optimize a
+ * custom Aggregate by providing a shortcut for when a single value is added to
+ * the input values.
+ *
+ * Some aggregation functions do not need to recalculate the aggregation of the
+ * whole set when one value changes. For example, when adding a new number to a
+ * series, the new sum of the series is the new value added to the previous sum.
+ *
+ * If it is not possible to shortcut the aggregation based on just one value
+ * being added, return `undefined` and the Metric will be completely
+ * recalculated.
+ *
+ * Where possible, if you are providing a custom Aggregate, seek an
+ * implementation of an AggregateAdd function that can reduce the complexity
+ * cost of growing the input data set. See the setMetricDefinition method for
+ * more examples.
+ *
+ * @category Aggregators
+ */
+export type AggregateAdd = (
+  metric: Metric,
+  add: number,
+  length: number,
+) => Metric | undefined;
+
+/**
+ * The AggregateRemove type describes a function that can be used to optimize a
+ * custom Aggregate by providing a shortcut for when a single value is removed
+ * from the input values.
+ *
+ * Some aggregation functions do not need to recalculate the aggregation of the
+ * whole set when one value changes. For example, when removing a number from a
+ * series, the new sum of the series is the new value subtracted from the
+ * previous sum.
+ *
+ * If it is not possible to shortcut the aggregation based on just one value
+ * being removed, return `undefined` and the Metric will be completely
+ * recalculated. One example might be if you were taking the minimum of the
+ * values, and the previous minimum is being removed. The whole of the rest of
+ * the list will need to be re-scanned to find a new minimum.
+ *
+ * Where possible, if you are providing a custom Aggregate, seek an
+ * implementation of an AggregateRemove function that can reduce the complexity
+ * cost of shrinking the input data set. See the setMetricDefinition method for
+ * more examples.
+ *
+ * @category Aggregators
+ */
+export type AggregateRemove = (
+  metric: Metric,
+  remove: number,
+  length: number,
+) => Metric | undefined;
+
+/**
+ * The AggregateReplace type describes a function that can be used to optimize a
+ * custom Aggregate by providing a shortcut for when a single value in the input
+ * values is replaced with another.
+ *
+ * Some aggregation functions do not need to recalculate the aggregation of the
+ * whole set when one value changes. For example, when replacing a number in a
+ * series, the new sum of the series is the previous sum, plus the new value,
+ * minus the old value.
+ *
+ * If it is not possible to shortcut the aggregation based on just one value
+ * changing, return `undefined` and the Metric will be completely
+ * recalculated.
+ *
+ * Where possible, if you are providing a custom Aggregate, seek an
+ * implementation of an AggregateReplace function that can reduce the complexity
+ * cost of changing the input data set in place. See the setMetricDefinition
+ * method for more examples.
+ *
+ * @category Aggregators
+ */
+export type AggregateReplace = (
+  metric: Metric,
+  add: number,
+  remove: number,
+  length: number,
+) => Metric | undefined;
+
+/**
+ * The MetricListener type describes a function that is used to listen to
+ * changes to a Metric.
+ *
+ * A MetricListener is provided when using the addMetricListener method. See
+ * that method for specific examples.
+ *
+ * When called, a MetricListener is given a reference to the Metrics object, the
+ * Id of the Metric that changed, and the new and old values of the Metric.
+ *
+ * If this is the first time that a Metric has had a value (such as when a table
+ * has gained its first row), the old value will be `undefined`. If a Metric now
+ * no longer has a value, the new value will be `undefined`.
+ *
+ * @category Listener
+ */
+export type MetricListener = (
+  metrics: Metrics,
+  metricId: Id,
+  newMetric: Metric | undefined,
+  oldMetric: Metric | undefined,
+) => void;
+
+/**
+ * The MetricsListenerStats type describes the number of listeners registered
+ * with the Metrics object, and can be used for debugging purposes.
+ *
+ * A MetricsListenerStats object is returned from the getListenerStats method,
+ * and is only populated in a debug build.
+ *
+ * @category Development
+ */
+export type MetricsListenerStats = {
+  metric?: number;
+};
+
+/**
+ * A Metrics object lets you define, query, and listen to, aggregations of Cell
+ * values within a Table in a Store.
+ *
+ * This is useful for counting the number of Row objects in a Table, averaging
+ * Cell values, or efficiently performing any arbitrary aggregations.
+ *
+ * Create a Metrics object easily with the createMetrics function. From there,
+ * you can add new Metric definitions (with the setMetricDefinition method),
+ * query their values (with the getMetric method), and add listeners for when
+ * they change (with the addMetricListener method).
+ *
+ * This module provides a number of predefined and self-explanatory aggregations
+ * ('sum', 'avg', 'min', and 'max'), and defaults to counting Row objects when
+ * using the setMetricDefinition method. However, far more complex aggregations
+ * can be configured with custom functions.
+ *
+ * @example
+ * This example shows a very simple lifecycle of a Metrics object: from
+ * creation, to adding a definition, getting an Metric, and then registering and
+ * removing a listener for it.
+ *
+ * ```tsx
+ * const store = createStore().setTable('species', {
+ *   dog: {price: 5},
+ *   cat: {price: 4},
+ *   worm: {price: 1},
+ * });
+ *
+ * const metrics = createMetrics(store);
+ * metrics.setMetricDefinition(
+ *   'highestPrice', // metricId
+ *   'species', //      tableId to aggregate
+ *   'max', //          aggregation
+ *   'price', //        cellId to aggregate
+ * );
+ *
+ * console.log(metrics.getMetric('highestPrice'));
+ * // -> 5
+ *
+ * const listenerId = metrics.addMetricListener('highestPrice', () => {
+ *   console.log(metrics.getMetric('highestPrice'));
+ * });
+ * store.setCell('species', 'horse', 'price', 20);
+ * // -> 20
+ *
+ * metrics.delListener(listenerId);
+ * metrics.destroy();
+ * ```
+ */
+export interface Metrics {
+  /**
+   * The setMetricDefinition method lets you set the definition of a Metric.
+   *
+   * Every Metric definition is identified by a unique Id, and if you re-use an
+   * existing Id with this method, the previous definition is overwritten.
+   *
+   * A Metric is an aggregation of numeric values produced from each Row within
+   * a single Table. Therefore the definition must specify the Table (by its Id)
+   * to be aggregated.
+   *
+   * Without the third `aggregate` parameter, the Metric will simply be a count
+   * of the number of Row objects in the Table. But often you will specify a
+   * more interesting aggregate - such as the four predefined aggregates, 'sum',
+   * 'avg', 'min', and 'max' - or a custom function that produces your own
+   * aggregation of an array of numbers.
+   *
+   * The fourth `getNumber` parameter specifies which Cell in each Row contains
+   * the numerical values to be used in the aggregation. Alternatively, a custom
+   * function can be provided that produces your own numeric value from the
+   * local Row as a whole.
+   *
+   * The final three parameters, `aggregateAdd`, `aggregateRemove`,
+   * `aggregateReplace` need only be provided when you are using your own custom
+   * `aggregate` function. These give you the opportunity to reduce your custom
+   * function's algorithmic complexity by providing shortcuts that can nudge an
+   * aggregation result when a single value is added, removed, or replaced in
+   * the input values.
+   *
+   * @param metricId The Id of the Metric to define.
+   * @param tableId The Id of the Table the Metric will be calculated from.
+   * @param aggregate Either a string representing one of a set of common
+   * aggregation techniques ('sum', 'avg', 'min', or 'max'), or a function that
+   * aggregates numeric values from each Row to create the Metric's overall
+   * value. Defaults to 'sum'.
+   * @param getNumber Either the Id of a Cell containing, or a function that
+   * produces, the numeric value that will be aggregated in the way specified by
+   * the `aggregate` parameter. Defaults to a function that returns `1` (meaning
+   * that if the `aggregate` and `getNumber` parameters are both omitted, the
+   * Metric will simply be a count of the Row objects in the Table).
+   * @param aggregateAdd A function that can be used to optimize a custom
+   * Aggregate by providing a shortcut for when a single value is added to the
+   * input values - for example, when a Row is added to the Table.
+   * @param aggregateRemove A function that can be used to optimize a custom
+   * Aggregate by providing a shortcut for when a single value is removed from
+   * the input values - for example ,when a Row is removed from the Table.
+   * @param aggregateReplace A function that can be used to optimize a custom
+   * Aggregate by providing a shortcut for when a single value in the input
+   * values is replaced with another - for example, when a Row is updated.
+   * @returns A reference to the Metrics object.
+   * @example
+   * This example creates a Store, creates a Metrics object, and defines a
+   * simple Metric to count the Row objects in the Table.
+   *
+   * ```tsx
+   * const store = createStore().setTable('species', {
+   *   dog: {price: 5},
+   *   cat: {price: 4},
+   *   worm: {price: 1},
+   * });
+   *
+   * const metrics = createMetrics(store);
+   * metrics.setMetricDefinition('speciesCount', 'species');
+   *
+   * console.log(metrics.getMetric('speciesCount'));
+   * // -> 3
+   * ```
+   * @example
+   * This example creates a Store, creates a Metrics object, and defines a
+   * standard Metric to get the highest value of each `price` Cell in the Row
+   * objects in the Table.
+   *
+   * ```tsx
+   * const store = createStore().setTable('species', {
+   *   dog: {price: 5},
+   *   cat: {price: 4},
+   *   worm: {price: 1},
+   * });
+   *
+   * const metrics = createMetrics(store);
+   * metrics.setMetricDefinition('highestPrice', 'species', 'max', 'price');
+   *
+   * console.log(metrics.getMetric('highestPrice'));
+   * // -> 5
+   * ```
+   * @example
+   * This example creates a Store, creates a Metrics object, and defines a
+   * custom Metric to get the lowest value of each `price` Cell, greater than 2.
+   *
+   * ```tsx
+   * const store = createStore().setTable('species', {
+   *   dog: {price: 5},
+   *   cat: {price: 4},
+   *   worm: {price: 1},
+   * });
+   *
+   * const metrics = createMetrics(store);
+   * metrics.setMetricDefinition(
+   *   'lowestPriceOver2',
+   *   'species',
+   *   (numbers) => Math.min(...numbers.filter((number) => number > 2)),
+   *   'price',
+   * );
+   *
+   * console.log(metrics.getMetric('lowestPriceOver2'));
+   * // -> 4
+   * ```
+   * @example
+   * This example also creates a Store, creates a Metrics object, and defines a
+   * custom Metric to get the lowest value of each `price` Cell, greater than 2.
+   * However, it also reduces algorithmic complexity with two shortcut
+   * functions.
+   *
+   * ```tsx
+   * const store = createStore().setTable('species', {
+   *   dog: {price: 5},
+   *   cat: {price: 4},
+   *   worm: {price: 1},
+   * });
+   *
+   * const metrics = createMetrics(store);
+   * metrics.setMetricDefinition(
+   *   'lowestPriceOver2',
+   *   'species',
+   *   (numbers) => Math.min(...numbers.filter((number) => number > 2)),
+   *   'price',
+   *   (metric, add) => (add > 2 ? Math.min(metric, add) : metric),
+   *   (metric, remove) => (remove == metric ? undefined : metric),
+   *   (metric, add, remove) =>
+   *     remove == metric
+   *       ? undefined
+   *       : add > 2
+   *       ? Math.min(metric, add)
+   *       : metric,
+   * );
+   *
+   * console.log(metrics.getMetric('lowestPriceOver2'));
+   * // -> 4
+   * store.setRow('species', 'fish', {price: 3});
+   * console.log(metrics.getMetric('lowestPriceOver2'));
+   * // -> 3
+   * ```
+   * @example
+   * This example creates a Store, creates a Metrics object, and defines a
+   * custom Metric to get the average value of a discounted price.
+   *
+   * ```tsx
+   * const store = createStore().setTable('species', {
+   *   dog: {price: 5, discount: 0.3},
+   *   cat: {price: 4, discount: 0.2},
+   *   worm: {price: 1, discount: 0.2},
+   * });
+   *
+   * const metrics = createMetrics(store);
+   * metrics.setMetricDefinition(
+   *   'averageDiscountedPrice',
+   *   'species',
+   *   'avg',
+   *   (getCell) => getCell('price') * (1 - getCell('discount')),
+   * );
+   *
+   * console.log(metrics.getMetric('averageDiscountedPrice'));
+   * // -> 2.5
+   * ```
+   * @category Configuration
+   */
+  setMetricDefinition(
+    metricId: Id,
+    tableId: Id,
+    aggregate?: 'sum' | 'avg' | 'min' | 'max' | Aggregate,
+    getNumber?: Id | ((getCell: GetCell, rowId: Id) => number),
+    aggregateAdd?: AggregateAdd,
+    aggregateRemove?: AggregateRemove,
+    aggregateReplace?: AggregateReplace,
+  ): Metrics;
+
+  /**
+   * The delMetricDefinition method removes an existing Metric definition.
+   *
+   * @param metricId The Id of the Metric to remove.
+   * @returns A reference to the Metrics object.
+   * @example
+   * This example creates a Store, creates a Metrics object, defines a simple
+   * Metric, and then removes it.
+   *
+   * ```tsx
+   * const store = createStore().setTable('species', {
+   *   dog: {price: 5},
+   *   cat: {price: 4},
+   *   worm: {price: 1},
+   * });
+   *
+   * const metrics = createMetrics(store);
+   * metrics.setMetricDefinition('speciesCount', 'species');
+   * console.log(metrics.getMetricIds());
+   * // -> ['speciesCount']
+   *
+   * metrics.delMetricDefinition('speciesCount');
+   * console.log(metrics.getMetricIds());
+   * // -> []
+   * ```
+   * @category Configuration
+   */
+  delMetricDefinition(metricId: Id): Metrics;
+
+  /**
+   * The getStore method returns a reference to the underlying Store that is
+   * backing this Metrics object.
+   *
+   * @returns A reference to the Store.
+   * @example
+   * This example creates a Metrics object against a newly-created Store and
+   * then gets its reference in order to update its data.
+   *
+   * ```tsx
+   * const metrics = createMetrics(createStore());
+   * metrics.setMetricDefinition('speciesCount', 'species');
+   * metrics.getStore().setCell('species', 'dog', 'price', 5);
+   * console.log(metrics.getMetric('speciesCount'));
+   * // -> 1
+   * ```
+   * @category Getter
+   */
+  getStore(): Store;
+
+  /**
+   * The getMetricIds method returns an array of the Metric Ids registered with
+   * this Metrics object.
+   *
+   * @returns An array of Ids.
+   * @example
+   * This example creates a Metrics object with two definitions, and then gets
+   * the Ids of the definitions.
+   *
+   * ```tsx
+   * const metrics = createMetrics(createStore())
+   *   .setMetricDefinition('speciesCount', 'species')
+   *   .setMetricDefinition('petsCount', 'pets');
+   *
+   * console.log(metrics.getMetricIds());
+   * // -> ['speciesCount', 'petsCount']
+   * ```
+   * @category Getter
+   */
+  getMetricIds(): Ids;
+
+  /**
+   * The getTableId method returns the Id of the underlying Table that is
+   * backing a Metric.
+   *
+   * If the Metric Id is invalid, the method returns `undefined`.
+   *
+   * @param metricId The Id of a Metric.
+   * @returns The Id of the Table backing the Metric, or `undefined`.
+   * @example
+   * This example creates a Metrics object, a single Metric definition, and then
+   * queries it (and a non-existent definition) to get the underlying Table Id.
+   *
+   * ```tsx
+   * const metrics = createMetrics(createStore());
+   * metrics.setMetricDefinition('speciesCount', 'species');
+   *
+   * console.log(metrics.getTableId('speciesCount'));
+   * // -> 'species'
+   * console.log(metrics.getTableId('petsCount'));
+   * // -> undefined
+   * ```
+   * @category Getter
+   */
+  getTableId(metricId: Id): Id | undefined;
+
+  /**
+   * The getMetric method gets the current value of a Metric.
+   *
+   * If the identified Metric does not exist (or if the definition references a
+   * Table or Cell value that does not exist) then `undefined` is returned.
+   *
+   * @param metricId The Id of the Metric.
+   * @returns The numeric value of the Metric, or `undefined`.
+   * @example
+   * This example creates a Store, creates a Metrics object, and defines a
+   * simple Metric to average the price values in the Table. It then uses
+   * getMetric to access its value (and also the value of a Metric that has not
+   * been defined).
+   *
+   * ```tsx
+   * const store = createStore().setTable('species', {
+   *   dog: {price: 5},
+   *   cat: {price: 4},
+   *   worm: {price: 1},
+   * });
+   *
+   * const metrics = createMetrics(store);
+   * metrics.setMetricDefinition('highestPrice', 'species', 'max', 'price');
+   *
+   * console.log(metrics.getMetric('highestPrice'));
+   * // -> 5
+   * console.log(metrics.getMetric('lowestPrice'));
+   * // -> undefined
+   * ```
+   * @category Getter
+   */
+  getMetric(metricId: Id): Metric | undefined;
+
+  /**
+   * The addMetricListener method registers a listener function with the Metrics
+   * object that will be called whenever the value of a specified Metric
+   * changes.
+   *
+   * You can either listen to a single Metric (by specifying the Metric Id as
+   * the method's first parameter), or changes to any Metric (by providing a
+   * `null` wildcard).
+   *
+   * The provided listener is a MetricListener function, and will be called with
+   * a reference to the Metrics object, the Id of the Metric that changed, the
+   * new Metric value, and the old Metric value.
+   *
+   * @param metricId The Id of the Metric to listen to, or `null` as a wildcard.
+   * @param listener The function that will be called whenever the Metric
+   * changes.
+   * @returns A unique Id for the listener that can later be used to remove it.
+   * @example
+   * This example creates a Store, a Metrics object, and then registers a
+   * listener that responds to any changes to a specific Metric.
+   *
+   * ```tsx
+   * const store = createStore().setTable('species', {
+   *   dog: {price: 5},
+   *   cat: {price: 4},
+   *   worm: {price: 1},
+   * });
+   *
+   * const metrics = createMetrics(store);
+   * metrics.setMetricDefinition('highestPrice', 'species', 'max', 'price');
+   *
+   * const listenerId = metrics.addMetricListener(
+   *   'highestPrice',
+   *   (metrics, metricId, newMetric, oldMetric) => {
+   *     console.log('highestPrice metric changed');
+   *     console.log([oldMetric, newMetric]);
+   *   },
+   * );
+   *
+   * store.setCell('species', 'horse', 'price', 20);
+   * // -> 'highestPrice metric changed'
+   * // -> [5, 20]
+   *
+   * metrics.delListener(listenerId);
+   * ```
+   * @example
+   * This example creates a Store, a Metrics object, and then registers a
+   * listener that responds to any changes to any Metric.
+   *
+   * ```tsx
+   * const store = createStore().setTable('species', {
+   *   dog: {price: 5},
+   *   cat: {price: 4},
+   *   worm: {price: 1},
+   * });
+   *
+   * const metrics = createMetrics(store)
+   *   .setMetricDefinition('highestPrice', 'species', 'max', 'price')
+   *   .setMetricDefinition('speciesCount', 'species');
+   *
+   * const listenerId = metrics.addMetricListener(
+   *   null,
+   *   (metrics, metricId, newMetric, oldMetric) => {
+   *     console.log(`${metricId} metric changed`);
+   *     console.log([oldMetric, newMetric]);
+   *   },
+   * );
+   *
+   * store.setCell('species', 'horse', 'price', 20);
+   * // -> 'highestPrice metric changed'
+   * // -> [5, 20]
+   * // -> 'speciesCount metric changed'
+   * // -> [3, 4]
+   *
+   * metrics.delListener(listenerId);
+   * ```
+   * @category Listener
+   */
+  addMetricListener(metricId: IdOrNull, listener: MetricListener): Id;
+
+  /**
+   * The delListener method removes a listener that was previously added to the
+   * Metrics object.
+   *
+   * Use the Id returned by the addMetricListener method. Note that the Metrics
+   * object may re-use this Id for future listeners added to it.
+   *
+   * @param listenerId The Id of the listener to remove.
+   * @returns A reference to the Metrics object.
+   * @example
+   * This example creates a Store, a Metrics object, registers a listener, and
+   * then removes it.
+   *
+   * ```tsx
+   * const store = createStore().setTable('species', {
+   *   dog: {price: 5},
+   *   cat: {price: 4},
+   *   worm: {price: 1},
+   * });
+   *
+   * const metrics = createMetrics(store);
+   * metrics.setMetricDefinition('highestPrice', 'species', 'max', 'price');
+   *
+   * const listenerId = metrics.addMetricListener(
+   *   'highestPrice',
+   *   (metrics, metricId, newMetric, oldMetric) => {
+   *     console.log('highestPrice metric changed');
+   *   },
+   * );
+   *
+   * store.setCell('species', 'horse', 'price', 20);
+   * // -> 'highestPrice metric changed'
+   *
+   * metrics.delListener(listenerId);
+   *
+   * store.setCell('species', 'giraffe', 'price', 50);
+   * // -> undefined
+   * // The listener is not called.
+   * ```
+   * @category Listener
+   */
+  delListener(listenerId: Id): Metrics;
+
+  /**
+   * The destroy method should be called when this Metrics object is no longer
+   * used.
+   *
+   * This guarantees that all of the listeners that the object registered with
+   * the underlying Store are removed and it can be correctly garbage collected.
+   *
+   * @example
+   * This example creates a Store, adds a Metrics object with a definition (that
+   * registers a RowListener with the underlying Store), and then destroys it
+   * again, removing the listener.
+   *
+   * ```tsx
+   * const store = createStore().setTable('species', {
+   *   dog: {price: 5},
+   *   cat: {price: 4},
+   *   worm: {price: 1},
+   * });
+   *
+   * const metrics = createMetrics(store);
+   * metrics.setMetricDefinition('speciesCount', 'species');
+   * console.log(store.getListenerStats().row);
+   * // -> 1
+   *
+   * metrics.destroy();
+   *
+   * console.log(store.getListenerStats().row);
+   * // -> 0
+   * ```
+   * @category Lifecycle
+   */
+  destroy(): void;
+
+  /**
+   * The getListenerStats method provides a set of statistics about the
+   * listeners registered with the Metrics object, and is used for debugging
+   * purposes.
+   *
+   * The statistics are only populated in a debug build: production builds
+   * return an empty object. The method is intended to be used during
+   * development to ensure your application is not leaking listener
+   * registrations, for example.
+   *
+   * @returns A MetricsListenerStats object containing Metrics listener
+   * statistics.
+   * @example
+   * This example gets the listener statistics of a Metrics object.
+   *
+   * ```tsx
+   * const store = createStore();
+   * const metrics = createMetrics(store);
+   * metrics.addMetricListener(null, () => console.log('Metric changed'));
+   *
+   * console.log(metrics.getListenerStats());
+   * // -> {metric: 1}
+   * ```
+   * @category Development
+   */
+  getListenerStats(): MetricsListenerStats;
+}
+
+/**
+ * The createMetrics function creates a Metrics object, and is the main entry
+ * point into the metrics module.
+ *
+ * It is trivially simple.
+ *
+ * A given Store can only have one Metrics object associated with it. If you
+ * call this function twice on the same Store, your second call will return a
+ * reference to the Metrics object created by the first.
+ *
+ * @param store The Store for which to register Metric definitions.
+ * @returns A reference to the new Metrics object.
+ * @example
+ * This example creates a Metrics object.
+ *
+ * ```tsx
+ * const store = createStore();
+ * const metrics = createMetrics(store);
+ * console.log(metrics.getMetricIds());
+ * // -> []
+ * ```
+ * @example
+ * This example creates a Metrics object, and calls the method a second time
+ * for the same Store to return the same object.
+ *
+ * ```tsx
+ * const store = createStore();
+ * const metrics1 = createMetrics(store);
+ * const metrics2 = createMetrics(store);
+ * console.log(metrics1 === metrics2);
+ * // -> true
+ * ```
+ */
+export function createMetrics(store: Store): Metrics;