UI libraries / Recommend / API reference / recommend-js

The trendingFacets function lets you render Trending Facets.

The Trending Products and Trending Facets models are a beta product.

Installation

The Recommend JavaScript package is available on the npm registry.

1
2
3
yarn add @algolia/recommend-js
# or
npm install @algolia/recommend-js

If you don’t want to use a package manager, you can use a standalone endpoint:

1
2
3
4
5
6
<script src="https://cdn.jsdelivr.net/npm/@algolia/recommend-js"></script>
<script src="https://cdn.jsdelivr.net/npm/@algolia/recommend"></script>
<script>
  const { frequentlyBoughtTogether, relatedProducts, trendingItems, trendingFacets } = window['@algolia/recommend-js'];
  const recommend = window['@algolia/recommend'];
</script>

Usage

To get started, you need to specify a container to inject the component into.

1
<div id="trendingFacets"></div>

Then, you can call the trendingFacets function and provide the container. It can be a CSS selector or an Element.

You can customize how to render each item by passing a custom component to the itemComponent prop.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
/** @jsx h */
import { h } from 'preact';
import { trendingFacets } from '@algolia/recommend-js';
import recommend from '@algolia/recommend';

const recommendClient = recommend('YourApplicationID', 'YourSearchOnlyAPIKey');
const indexName = 'YOUR_INDEX_NAME';
const facetName = 'YOUR_FACET_NAME';

trendingFacets({
  container: '#trendingFacets',
  recommendClient,
  indexName,
  facetName,
  itemComponent({ item }) {
    return (
      <pre>
        <code>{JSON.stringify(item)}</code>
      </pre>
    );
  },
});

Horizontal slider view

You can also wrap the component within a custom view. For example, you can use the provided HorizontalSlider UI component to display items as a scrollable slider.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
/** @jsx h */
import { h } from 'preact';
import { trendingFacets } from '@algolia/recommend-js';
import { horizontalSlider } from '@algolia/ui-components-horizontal-slider-js';
import recommend from '@algolia/recommend';

import '@algolia/ui-components-horizontal-slider-theme';

const recommendClient = recommend('YourApplicationID', 'YourSearchOnlyAPIKey');
const indexName = 'YOUR_INDEX_NAME';
const facetName = 'YOUR_FACET_NAME';

trendingFacets({
  container: '#trendingFacets',
  recommendClient,
  indexName,
  facetName,
  view: horizontalSlider,
  itemComponent({ item }) {
    return (
      <pre>
        <code>{JSON.stringify(item)}</code>
      </pre>
    );
  },
});

If you don’t use JSX in your project, you can return tagged templates using htm or leverage createElement and Fragment from itemComponent, headerComponent and fallbackComponent.

With createElement

You can use createElement and Fragment in every component props to create components without JSX. They’re bound to Preact 10’s createElement and Fragment.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
import { trendingFacets } from '@algolia/recommend-js';
import recommend from '@algolia/recommend';

const recommendClient = recommend('YourApplicationID', 'YourSearchOnlyAPIKey');
const indexName = 'YOUR_INDEX_NAME';
const facetName = 'YOUR_FACET_NAME';

trendingFacets({
  container: '#trendingFacets',
  recommendClient,
  facetName,
  indexName,
  itemComponent({ item, createElement, Fragment }) {
    return createElement(
      'pre',
      null,
      createElement('code', null, JSON.stringify(item))
    );
  },
});

With html

You can use the exposed html function. This function is bind with htm. It allows to pass templates as HTML strings.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
import { trendingFacets } from '@algolia/recommend-js';
import recommend from '@algolia/recommend';

const recommendClient = recommend('YourApplicationID', 'YourSearchOnlyAPIKey');
const indexName = 'YOUR_INDEX_NAME';
const currentObjectID = 'YOUR_OBJECT_ID';

trendingFacets({
  container: '#trendingFacets',
  recommendClient,
  indexName,
  objectIDs: [currentObjectID],
  itemComponent({ item, html }) {
    return html`<pre>
      <code>${JSON.stringify(item)}</code>
    </pre>`;
  },
});

Parameters

itemComponent
type: ({ item, createElement, Fragment, html }) => JSX.Element | VNode | VNode[]
Required

The function to display each item.

container
type: string | HTMLElement

The container for the component. You can either pass a CSS selector or an Element. If there are several containers matching the selector, it picks up the first one.

When undefined, the function returns a JSX element for you to inject wherever you want.

headerComponent
type: ({ classNames, recommendations, translations, Fragment, createElement, html }) => JSX.Element | VNode | VNode[]

The function to render a header for your items.

The default implementation is:

1
2
3
4
5
6
7
8
9
10
11
12
13
function HeaderComponent(props) {
  return (
    if (!props.translations.title) {
      return null;
    }

    return (
      <h3 className={cx('auc-Recommend-title', props.classNames.title)}>
        {props.translations.title}
      </h3>
    );
  );
}
fallbackComponent
type: ({ createElement, Fragment, html }) => JSX.Element | VNode | VNode[]

A fallback component to render when no recommendations are returned.

view
type: ViewProps

The view component to render your items into. For example, you can use the horizontalSlider UI component.

The horizontalSlider UI component is currently not supported with recommend-js when imported as a UMD bundle, only when using a package manager.

The default implementation is:

1
2
3
4
5
6
7
8
9
10
11
12
13
function ListView(props) {
  return (
    <div className="auc-Recommend-container">
      <ol className="auc-Recommend-list">
        {props.items.map(item => (
          <li key={item.objectID} className="auc-Recommend-item">
            <props.itemComponent item={item} />
          </li>
        ))}
      </ol>
    </div>
  );
}
environment
type: typeof window
default: window

The environment in which your application is running.

This is useful if you’re using Recommend in a different context than window.

This function also accepts all the props that useTrendingFacets supports:

recommendClient
type: RecommendClient
Required

The initialized Algolia recommend client.

indexName
type: string
Required

The name of the target index.

maxRecommendations
type: number

The number of recommendations to retrieve. Depending on the available recommendations and the other request parameters, the actual number of hits may be lower than that. If maxRecommendations isn’t provided or set to 0, all matching recommendations are returned, and no fallback request is performed.

threshold
type: number

Threshold for the recommendations confidence score (between 0 and 100). Only recommendations with a greater score are returned.

transformItems
type: (items: Array<RecordWithObjectID<TItem>>) => items

A function to transform the retrieved items before passing them to the component. It’s useful to add or remove items, change them, or reorder them.

facetName
type: string
Required

The facet attribute to get recommendations for.

Did you find this page helpful?