dnalob / pivot-chart

light and fast implementation of web pivot table / pivot chart components.

English | 简体中文

Pivot-Chart

pivot chart is a an extension data visualization type of pivot table. It allows user to observe the data in different chart type without limited to table and pure numbers.

pivot chart also provide with basic pivot table components for you to build your web apps, you can regard pivot table as a member in the subset of pivot charts. pivot chart is build based on cube-core: an MOLAP cube solution in js.

Demo

Online Demo

feature	demo(gif)
basic expandable nest/cross table
custom aggregator of measures
different visualization type

Usage

install npm package.

npm i --save pivot-chart

# or

yarn add pivot-chart

basic usage.

import { PivotChart } from 'fast-pivot';

function App () {
  return <PivotChart
    visType={visType}
    dataSource={data}
    rows={rows}
    columns={columns}
    measures={measures} 
    />
}

demo above can be run locally with

# init development environment
yarn workspace react-pivot-table initenv
# start dev server
yarn workspace react-pivot-table dev

API

Types

Type	Desc
Field	- id  - name  - aggName  aggregator's name. - cmp <(a: any, b: any) => number>
Measure	extends Field - aggregator aggregator function. - minWidth  - formatter <value: number> => number
VisType	currently support number , bar , line , scatter .
Record	a plain javascript object
DataSource	Record[] , Array of Record.
QueryNode	- dimCode  - dimValue
QueryPath	<QueryNode[]>. example: [{dimCode: 'Sex', dimValue: 'male'}, {dimCode: 'Age', dimValue:'*'}]

common

common interface are those used in both async pivot table and sync pivot table.

• rows: Field[] dimensions in row. required
• columns: Field[] dimensions in column.required
• measures: Measure[] measures or indicators.required
• visType: VisType mark type in cell. number as default(which is a common pivot table).optional
• defaultExpandedDepth. default expanded level in row or column nest tree.optional
  • defaultExpandedDepth.rowDepth: number 
  • defaultExpandedDepth.columnDepth: number 
• showAggregatedNode: {row: boolean; column: boolean} whether display aggregated node for each group. optional

SyncPivotTable

• dataSource: Record[] . Array of record(normaly plain object). (json style data array). required

example:

import React, { useEffect, useState, useMemo } from 'react';
import ReactDOM from 'react-dom';
import { getTitanicData } from './mock';
import { ToolBar, PivotChart, DragableFields, Aggregators, DataSource, VisType, DraggableFieldState } from '../src/index';

const { dataSource, dimensions, measures } = getTitanicData();
const fields = dimensions.concat(measures).map(f => ({ id: f, name: f }));

const initDraggableState: DraggableFieldState = {
  fields: [],
  rows: [],
  columns: [],
  measures: []
};

function App () {
  const [data, setData] = useState<DataSource>([]);
  const [fstate, setFstate] = useState<DraggableFieldState>(initDraggableState)
  const [visType, setVisType] = useState<VisType>('number');
  useEffect(() => {
    setData(dataSource);
  }, [])
  const measures = useMemo(() => fstate['measures'].map(f => ({
    ...f,
    aggregator: Aggregators[(f.aggName || 'sum') as keyof typeof Aggregators]
  })), [fstate['measures']]);
  return <div>
    <DragableFields onStateChange={(state) => {setFstate(state)}} fields={fields} />
    <ToolBar visType={visType} onVisTypeChange={(type) => { setVisType(type) }} />
    <PivotChart visType={visType} dataSource={data} rows={fstate['rows']} columns={fstate['columns']} measures={measures} />
  </div>
}

ReactDOM.render(<App />, document.getElementById('root'))

AsyncPivotTable

• cubeQuery: (path: QueryPath, measures: string[]) => Promise; . A function handle for cube query. path is the groupby dimension path(made of a series of dimension and its value). measures are the fields needed to be aggregated. required
• branchFilter: bad api, not recommanded to use it. a fake filter whihc only control display of node and not influence the aggregated result of parent node. optional
• dimensionCompare: (a: string, b: string) => number .compare function for sort of dimension node. optional

function AsyncApp () {
  
  const [data, setData] = useState<DataSource>([]);
  const [fields, setFields] = useState<Field[]>([]);
  const [fstate, setFstate] = useState<DraggableFieldState>(initDraggableState)
  const [visType, setVisType] = useState<VisType>('number');
  useEffect(() => {
    const { dataSource, dimensions, measures } = getTitanicData();
    setData(dataSource);
    const fs: Field[] = [...dimensions, ...measures].map((f: string) => ({ id: f, name: f }));
    setFields(fs);
  }, [])
  const measures = useMemo(() => fstate['measures'].map(f => ({
    ...f,
    aggregator: Aggregators[(f.aggName || 'sum') as keyof typeof Aggregators],
    minWidth: 100,
    formatter: f.id === 'Survived' && ((val: any) => `${val} *`)
  })), [fstate['measures']]);
  const cubeQuery = useCallback(async (path: QueryPath, measures: string[]) => {
    return TitanicCubeService(path.map(p => p.dimCode), measures);
  }, [])
  return <div>
    <DragableFields onStateChange={(state) => {setFstate(state)}} fields={fields} />
    <ToolBar visType={visType} onVisTypeChange={(type) => { setVisType(type) }} />
    <AsyncPivotChart
      visType={visType}
      rows={fstate['rows']}
      columns={fstate['columns']}
      async
      defaultExpandedDepth={{
        rowDepth: 20,
        columnDepth: 20
      }}
      cubeQuery={cubeQuery}
      measures={measures} />
  </div>
}

Theme

• Theme.registerTheme(theme: ThemeConfig)
• ThemeConfig 

interface ThemeConfig {
  root?: {
    display?: boolean,
    label?: string
  },
  summary?: {
    label?: string
  },
  table?: {
    thead?: {
      backgroundColor?: string;
      color?: string;
    }
    borderColor?: string;
    color?: string;
  }
}
// default config
const THEME_CONFIG: ThemeConfig = {
  root: {
    display: true,
    label: 'All'
  },
  summary: {
    label: '(total)'
  },
  table: {
    thead: {
      backgroundColor: '#E9EDF2',
      color: '#5A6C84'
    },
    borderColor: '#DFE3E8',
    color: '#333333'
  }
};

import React, { useEffect, useState, useMemo } from 'react';
import ReactDOM from 'react-dom';
import { getTitanicData } from './mock';
import { ToolBar, PivotChart, DragableFields, Aggregators, DataSource, VisType, DraggableFieldState, Theme } from '../src/index';

const { dataSource, dimensions, measures } = getTitanicData();
const fields = dimensions.concat(measures).map(f => ({ id: f, name: f }));

const initDraggableState: DraggableFieldState = {
  fields: [],
  rows: [],
  columns: [],
  measures: []
};

Theme.registerTheme({
  root: {
    display: true,
    label: 'root'
  },
  summary: {
    label: '(total)'
  }
})

function App () {
  const [data, setData] = useState<DataSource>([]);
  const [fstate, setFstate] = useState<DraggableFieldState>(initDraggableState)
  const [visType, setVisType] = useState<VisType>('number');
  useEffect(() => {
    setData(dataSource);
  }, [])
  const measures = useMemo(() => fstate['measures'].map(f => ({
    ...f,
    aggregator: Aggregators[(f.aggName || 'sum') as keyof typeof Aggregators]
  })), [fstate['measures']]);
  return <div>
    <DragableFields onStateChange={(state) => {setFstate(state)}} fields={fields} />
    <ToolBar visType={visType} onVisTypeChange={(type) => { setVisType(type) }} />
    <PivotChart visType={visType} dataSource={data} rows={fstate['rows']} columns={fstate['columns']} measures={measures} />
  </div>
}

ReactDOM.render(<App />, document.getElementById('root'))

Others

another implementation of cube-core pivot-table can be found as ./packages/demo

Common Question

SyncPivotChart vs. AsyncPivotChart ?

Sync Pivot Chart does all cube computaion in frontend.(In future, it may do those work in webworker and it will seems to be async). Those cube query can be speeded up by cube-core which can reuse similarly groupby result.
Async Pivot Chart can use cube query from server or customed implementation(either on server or browser, async or sync), but developer need to figure out how to speed up those by themsleves.