本文是对一本书中有关Chart注意事项的摘抄，如下：

  • 什么是Graph应用的成功？
    • The important criterion for a graph is not simply how fast we can see a result; rather, it is whether through the use of the graph we can see something that … could not have been seen at all. If a graphical display requires hours of study to make a discovery that would have gone undetected without the graph, then the display is a success

 

  • Graph功能的基本定义：
    • Graph主要为了实现随机数据的显示，并不是指那些用手工精心绘制的Graph
    • Graph是为了帮助用户解决问题的，所以必须允许用户熟练的操作Graph，这些操作包括：改变数据、行列互换、或者修改Title，Color和Style
    • Graph应该可以在各种不同的类型之间自由转换，但若这种自由转换会带来潜在的问题时，系统应该能够帮助用户防止这些问题的发生
    • 用户应该可以在Graph的基础上作深入的操作，包括Drill down To other Graph or data，构建查询，协同显示等功能
    • 最重要的是，Graph不仅仅是帮助用户快捷直观的察看结果的工具，而且是分析和解决复杂问题的有力工具

 

  • 何时使用Chart？
    • Less than or equal to 3 numbers: Use a sentence.
    • 4 to 20 numbers: Use a table.
    • More than 20 numbers: Use a graph.

 

  • 使用Chart背景线的注意事项
    • Chart一定要比背景线明显：The grids go in the background. Don’t put the grid over the data. Grids are always background, not foreground; the data must be more obvious than the grid.
    • 允许用户开关背景线
    • 让比较更加容易
      • 应该为用户提供Graph比较功能，比较时，应该能沟让用户同时看到两个或者两个以上的得Graph。
      • 应通过背景线为用户提供良好的比较功能，需要注意：use the same grid at the same scale in all graphs in the set.
  • 使用轴的注意事项
    • X，Y轴的基本作用
      • The horizontal, or X, axis shows the time or cause of an event—the independent variable
      • The vertical, or Y, axis shows the caused effect—the dependent variable.
    • Size the Data Rectangle Correctly
      • 数据区域不能太大，若数据的最大值是600，而数据区域达到1200，就会出现大量空白
      • 同时，为了实现在两个数据区域之间的比较功能，两个数据区域应该使用同样的Scale，即使Chart控件可以根据数值大小自动调整Scale，也应该强制使这些需要进行比较的Chart的Scale相同
    • Square Up the Data
      • 若行上是Days，而列上的数值是100的倍数，则这两个轴上数据点的间距应该相同，以使背景线成为正方形
      • 同时，要提供给用户改变XY轴长宽比率的功能，以便于用户通过改变行列比率显示数据中细节的趋势
  • 如何使用标尺上的标注
    • 标注的分类
      • 类别标注：A category scale (also called qualitative or nominal) consists of an ordered or unordered series of words or numbers that identify people, places, things, etc. An example would be “Stocks, Bonds, Funds, Cash.”
      • 数量标注：A quantitative scale (also called value, interval, numeric, or amount) consists of numbers in sequence with meaningful and uniform spacing between them. An example would be “$100, $200, $300, $400.”
      • 序列标注：A sequence scale consists of words or numbers in an ordered sequence with uniform spacing between them. Time-series and orderof- occurrence scales are widely used sequence scales (often on the independent axis). An example would be “Jan, Feb,Mar, Apr,May.”
    • 使用标注的注意事项
      • 要有起始点，一般为0，0标注，若数据范围不包括0，0 点，则使用最小值标注起始点。同样，结束点最好也要有标示
      • 对于数据标注，使用1，2或者5的倍数作为数据的间隔，如下：
        • Intervals of 1: 0, 1, 2, 3 …
        • Intervals of 20: 0, 20, 40, 60, 80 …
        • Intervals of 0.5: 0, 0.05, 0.10, 0.15, 0.20 …
      • 对于序列标示，For sequence or time-series scales, customary intervals include hours of the day, days of the week, and weeks or months of the year.
      • 当标注中有小数时，不要省略0，如，.1要写为0.1，因为.1的.太容易被忽略
      • 若数据点恰好在X轴上，则将0值轴进行偏移，使之与X轴的标注分离开，但要特别注意，进行偏移后一定要在X轴和0轴之间增加一些空隙，以提醒用户
    • 使用标注标签的注意事项
      • 不要使用太多标签：The purpose of tick marks is not so much to identify every data point but rather to identify the type of scale.
      • 标签要放在轴的外边
  • 如何使用Label
    • Label的分类
      • Title, centered at top (or flush left if it would overlap a key at the right).
      • Y-axis label, either centered above the Y-axis scale or turned sideways and centered facing the Y-axis scale.
      • X-axis label, centered below the X-axis scale. Keys (also called legends).
      • Data labels, which hold the names or values for individual data points.
      • Tick or scale labels, which are described in “Don’t Use Too Many Tick Marks” above.
    • Title的使用注意事项
      • The title should use a larger font than any other label on the graph and be centered at the top of the frame.
      • 提供默认的Title供用户修改
    • 轴Label的注意事项
      • 描述度量的单位
      • 标明Data Source，the name of the table in the database, the file name and location, the server name, or whatever will be most useful. 可以通过在X Label下以小字或者Caption的方式来标明当前数据的来源
      • Y轴标签不要使用分散字母竖排的显示，应该将整个单词竖排
      • 拼出所有单词，若一定要使用缩写，也必须使用标准的缩写
  • Chart类型的选择（Using Based Chart Type）
    • 主要的使用目的分类
      • 简单对比：Simple comparisons.
      • 时间变化趋势：Changes over time.
      • 统计分析：Statistical analysis.
      • 比例：Proportion.
    • 适用用于简单对比的Chart Type以及注意事项
      • Bar Chart
        • Bar之间要有空隙，空隙的大小是Bar宽度的1.5倍为宜：The spacing between bars or sets of bars should be one-half the size of the bars.
      • Horizontal Bar Chart
        • 适合于解决较长标注Label的问题，可以将Label放到右边
      • Clustered Bar Chart
        • 适合用来比较2到4个系列的数据
      • Zero-Line Bar Chart
        • 适合有负值的情况
    • 适合用于时间变化趋势的Chart Type
      • Line Chart
        • 适用于表现趋势和两个值之间的对比
        • 拐角和标注点说明Line Chart由一系列实际值组成
        • 光滑曲线说明Line Chart中使用了插值
        • 虚线或者灰色的线说明在使用推测值
      • K线：股票和金融工具，略
    • 统计分析Chart：略
    • 适合于比例分析的Chart
      • Area Chart
        • 即可表现每一个值系列有可以表现总体趋势
        • 需要将变化比较平缓的Area放到最下面，否则一个在底部剧烈变化的系列会使所有系列看起来都有剧烈的变化
        • 注意Line Chart和Area Chart之间转换的问题
          • Do not confuse area charts with line graphs. Although Excel and other graphing programs let users change the areas between the lines in line graphs into filled areas, the filled areas have no meaning. In area charts, the filled areas are actually volumes.
          • If area charts rarely appear in the domain for which you’re designing graphs, either avoid them or let users transform them into pie charts or segmented bar graphs. Most people understand pie charts and segmented bar graphs more quickly.
      • Pie Chart
        • 尽量不要使用多个Pie Chart进行系列值对比工作，因为非常难以阅读，应该使用其它的Chart类别进行比较
          • It is not possible to compare two or more data series without showing multiple pie charts. However, most people find it hard to compare wedge-shaped areas from one pie chart to the next. If you need to compare data series, use a different type of graph—area charts, segmented bars, and donut charts (shown later) let readers compare multiple series.
        • 格式化Pie Chart时的注意事项
          • 按照顺序排列Pie Chart的分类：Pie Chart应该从12点方向开始，按次序依次排放分类，即不应该将最大的放到前面，也不应该将排名第一的放到前面
          • 如果可能，尽量不要使用3D的Pie Chart表现数据，因为不同的视角会影响读者对比例的判断
      • Donut Chart Variation
        • 图略
      • stacked bar charts
        • are good for showing proportional relationships (like pie charts and area charts) over time (like bar charts).
        • They can be transformed easily into area charts.
      • Zero Line Bar Charts
        • 图略