搜索

PythonReactSEOPDF
GitHub Skill

chartjs-chart-types

This skill should be used when the user asks "how to create a line chart", "Chart.js bar chart", "pie chart Chart.js", "doughnut chart", "radar chart", "polar area chart", "bubble chart", "scatter chart", "mixed chart", "combo chart", "area chart", "stacked chart", "horizontal bar chart", "Chart.js chart types", "dataset properties", "chart data structure", or needs help implementing specific Chart.js v4.5.1 chart types.

sjnims☆ 0 仓库 Stars0 仓库 Forks

安装与使用

请根据来源仓库和你的 AI 工具说明安装此技能。先阅读技能文件和依赖要求,再添加到工作流。

查看来源 ↗使用指南

技能说明

Chart.js Chart Types (v4.5.1)

Complete guide to all Chart.js chart types, their configuration, and dataset properties.

Available Chart Types

Type Description Use Case
line Line chart Trends over time
bar Bar chart Comparing categories
pie Pie chart Parts of a whole
doughnut Doughnut chart Parts of a whole with center
radar Radar/spider chart Multivariate data comparison
polarArea Polar area chart Cyclical data
bubble Bubble chart Three dimensions
scatter Scatter plot Correlation analysis

Line Chart

Show trends and changes over time:

new Chart(ctx, {
  type: 'line',
  data: {
    labels: ['Jan', 'Feb', 'Mar', 'Apr', 'May'],
    datasets: [{
      label: 'Sales',
      data: [65, 59, 80, 81, 56],
      fill: false,
      borderColor: 'rgb(75, 192, 192)',
      tension: 0.1  // 0 = straight lines, 0.4 = curved
    }]
  }
});
Key Line Properties
Property Type Description
tension number Bezier curve tension (0-1)
fill boolean/string Fill area under line
stepped boolean/string Stepped line interpolation
borderDash number[] Dashed line pattern
pointRadius number Data point size
pointStyle string Point shape (circle, rect, triangle, etc.)
spanGaps boolean Connect over null values
Vertical Line Chart

Flip axes with indexAxis:

options: {
  indexAxis: 'y'  // Vertical line chart
}

Bar Chart

Compare discrete categories:

new Chart(ctx, {
  type: 'bar',
  data: {
    labels: ['Red', 'Blue', 'Yellow', 'Green'],
    datasets: [{
      label: 'Votes',
      data: [12, 19, 3, 5],
      backgroundColor: [
        'rgba(255, 99, 132, 0.5)',
        'rgba(54, 162, 235, 0.5)',
        'rgba(255, 206, 86, 0.5)',
        'rgba(75, 192, 192, 0.5)'
      ],
      borderWidth: 1
    }]
  },
  options: {
    scales: {
      y: { beginAtZero: true }
    }
  }
});
Key Bar Properties
Property Type Description
barThickness number/'flex' Bar width in pixels
barPercentage number Bar width relative to category (0-1)
categoryPercentage number Category width (0-1)
borderRadius number/object Rounded corners
borderSkipped string Which border to skip
Horizontal Bar Chart
options: {
  indexAxis: 'y'  // Makes bars horizontal
}
Stacked Bar Chart
options: {
  scales: {
    x: { stacked: true },
    y: { stacked: true }
  }
}

Pie & Doughnut Charts

Show proportional data:

// Pie chart
new Chart(ctx, {
  type: 'pie',
  data: {
    labels: ['Red', 'Blue', 'Yellow'],
    datasets: [{
      data: [300, 50, 100],
      backgroundColor: ['#ff6384', '#36a2eb', '#ffce56']
    }]
  }
});

// Doughnut chart
new Chart(ctx, {
  type: 'doughnut',
  data: { /* same as pie */ },
  options: {
    cutout: '50%'  // Size of center hole
  }
});
Key Pie/Doughnut Properties
Property Type Description
cutout number/string Center hole size (doughnut)
rotation number Starting angle (degrees)
circumference number Arc sweep (degrees)
offset number Arc offset on hover
hoverOffset number Offset when hovering
Semi-Circle Chart
options: {
  rotation: -90,
  circumference: 180
}

Radar Chart

Compare multiple variables across categories:

new Chart(ctx, {
  type: 'radar',
  data: {
    labels: ['Speed', 'Power', 'Defense', 'Stamina', 'Agility'],
    datasets: [{
      label: 'Player A',
      data: [65, 59, 90, 81, 56],
      fill: true,
      backgroundColor: 'rgba(255, 99, 132, 0.2)',
      borderColor: 'rgb(255, 99, 132)',
      pointBackgroundColor: 'rgb(255, 99, 132)'
    }, {
      label: 'Player B',
      data: [28, 48, 40, 19, 96],
      fill: true,
      backgroundColor: 'rgba(54, 162, 235, 0.2)',
      borderColor: 'rgb(54, 162, 235)'
    }]
  },
  options: {
    scales: {
      r: {
        angleLines: { display: true },
        suggestedMin: 0,
        suggestedMax: 100
      }
    }
  }
});

Polar Area Chart

Like pie but with equal angles, varying radius:

new Chart(ctx, {
  type: 'polarArea',
  data: {
    labels: ['Red', 'Green', 'Yellow', 'Grey', 'Blue'],
    datasets: [{
      data: [11, 16, 7, 3, 14],
      backgroundColor: [
        'rgba(255, 99, 132, 0.5)',
        'rgba(75, 192, 192, 0.5)',
        'rgba(255, 205, 86, 0.5)',
        'rgba(201, 203, 207, 0.5)',
        'rgba(54, 162, 235, 0.5)'
      ]
    }]
  }
});

Scatter Chart

Show correlation between two variables:

new Chart(ctx, {
  type: 'scatter',
  data: {
    datasets: [{
      label: 'Data Points',
      data: [
        { x: 10, y: 20 },
        { x: 15, y: 10 },
        { x: 20, y: 30 },
        { x: 25, y: 25 }
      ],
      backgroundColor: 'rgb(255, 99, 132)'
    }]
  },
  options: {
    scales: {
      x: { type: 'linear', position: 'bottom' },
      y: { type: 'linear' }
    }
  }
});

Bubble Chart

Three-dimensional data visualization:

new Chart(ctx, {
  type: 'bubble',
  data: {
    datasets: [{
      label: 'Dataset',
      data: [
        { x: 20, y: 30, r: 15 },  // r = bubble radius
        { x: 40, y: 10, r: 10 },
        { x: 30, y: 22, r: 25 }
      ],
      backgroundColor: 'rgba(255, 99, 132, 0.5)'
    }]
  },
  options: {
    aspectRatio: 1  // Square chart for bubbles
  }
});

Area Chart

Line chart with filled area:

new Chart(ctx, {
  type: 'line',
  data: {
    labels: ['Jan', 'Feb', 'Mar', 'Apr', 'May'],
    datasets: [{
      label: 'Revenue',
      data: [65, 59, 80, 81, 56],
      fill: true,  // Enable fill
      backgroundColor: 'rgba(75, 192, 192, 0.2)',
      borderColor: 'rgb(75, 192, 192)'
    }]
  }
});
Fill Options
Value Description
false No fill (default)
true Fill to origin
'origin' Fill to origin
'start' Fill to bottom
'end' Fill to top
'-1' Fill to previous dataset
'+1' Fill to next dataset

Mixed Charts

Combine multiple chart types:

new Chart(ctx, {
  type: 'bar',  // Base type
  data: {
    labels: ['Jan', 'Feb', 'Mar', 'Apr'],
    datasets: [
      {
        type: 'line',  // Override type for this dataset
        label: 'Trend',
        data: [50, 60, 70, 80],
        borderColor: 'rgb(255, 99, 132)',
        fill: false
      },
      {
        type: 'bar',  // Explicit bar type
        label: 'Sales',
        data: [40, 55, 65, 75],
        backgroundColor: 'rgba(54, 162, 235, 0.5)'
      }
    ]
  }
});
Drawing Order

Control which dataset renders on top:

datasets: [
  { label: 'Bars', order: 2 },  // Drawn first (behind)
  { label: 'Line', order: 1 }   // Drawn last (on top)
]

Data Structures

Standard Format
data: {
  labels: ['A', 'B', 'C'],
  datasets: [{
    data: [10, 20, 30]
  }]
}
Object Format
datasets: [{
  data: [
    { x: 'A', y: 10 },
    { x: 'B', y: 20 },
    { x: 'C', y: 30 }
  ]
}]
Parsed Data Format
datasets: [{
  data: [
    { id: 'Sales', value: 100 },
    { id: 'Revenue', value: 200 }
  ],
  parsing: {
    xAxisKey: 'id',
    yAxisKey: 'value'
  }
}]

Additional Resources

  • See references/dataset-properties.md for complete property reference
  • See examples/ for working HTML examples:
    • line-area-charts.html - Line charts with tension, area fills, stepped lines
    • bar-charts.html - Vertical, horizontal, stacked bars, and mixed bar+line
    • circular-charts.html - Pie, doughnut, semi-circle gauge, polar area, radar

使用示例

查看来源仓库中的使用示例与配置说明。

查看来源 ↗

来源历史

收录时间
2026-01-20
索引更新时间
2026-01-26
Branch
main

收录与同步日期属于本站记录。源码修改历史请查看来源仓库。

GitHub ↗

chat Comments (0)

chat_bubble_outline

No comments yet. Be the first to share your thoughts!