ae制作数据可视化

by Krist Wongsuphasawat

克里斯特·旺苏帕萨瓦(Krist Wongsuphasawat)

(How I carefully crafted a truly terrible data visualization)

Yes, you read that right. I am going to explain how I put together a really bad visualization, intentionally.

是的，你看的没错。 我将解释如何有意地将一个非常糟糕的可视化组合在一起。

Andy Kirk of visualisingdata.com posted an interesting contest challenging everyone to come up with the “best worst viz.” Of course, one of the motivations for me doing this is to win a copy of his book. But the contest itself is also a thoughtful exercise.

visualisingdata.com的安迪·科克(Andy Kirk)发布了一个有趣的竞赛，要求所有人提出“最好的最糟糕的结果”。 当然，我这样做的动机之一就是赢得他的书的副本。 但是比赛本身也是一项深思熟虑的活动。

When talking about extremely bad visualizations, the stereotypical ones often involve 3D pie charts, rainbow color palettes and terrible choices of fonts, layouts, and colors.

当谈到极差的可视化时，陈规定型的通常涉及3D饼图， 彩虹色调色板和 字体，布局和颜色的糟糕选择。

In my opinion, bad visualizations don’t have to be just that. The goal I had in mind was to create a piece that looks totally harmless, but will torture your brain until you realize how absurdly ridiculous the whole thing is.

在我看来，糟糕的可视化并不仅限于此。 我想到的目标是创建一块看上去完全无害的作品，但是会折磨您的大脑，直到您意识到整件事情多么荒谬可笑。

I collected data from visualizations featured on viz.wtf and drew each mark to represent one of the visualizations and its properties. Example questions to exercise your WTF gland are:

我从viz.wtf上的可视化文件中收集了数据，并绘制了每个标记以表示可视化文件及其属性之一。 行使WTF腺的示例问题包括：

  • What’s the most common color for these visualizations?
  • Where are the pie charts?
  • Can you point out the least popular piece?
  • How often is 3D used?
  • Is there any pattern at all?

Before reading the next section, try to figure out everything that is wrong with this chart by yourself.

在阅读下一部分之前，请尝试自行找出该图表的所有错误。

(Concept)

The main idea was to create conflicts in perception and mess with viewers’ cognitive thinking.

主要思想是在感知上产生冲突，并与观看者的认知思维陷入混乱。

Bad visualizations usually have mismatches between visual encodings and data, such as encoding incomparable areas (3D pie) for numerical values. These mismatches leave viewers with little to do but scratch their heads, then abandon the visualization because it takes too much effort to make sense of it.

不良的可视化效果通常在视觉编码和数据之间不匹配，例如对数值进行无可比拟的区域编码(3D饼图)。 这些不匹配使得观看者几乎无所事事，但会挠头，然后放弃可视化，因为要花费太多精力才能理解它。

I wanted to take bad to the next level, and was inspired by one of my favorite responses from the Stack Overflow questions, “What is the best comment in source code you have ever encountered?”

我想把自己提升到一个新的水平，并受到Stack Overflow问题中我最喜欢的回答之一的启发，“ 您遇到过的源代码中最好的注释是什么？ ”

#define TRUE FALSE

＃定义真假

My goal was to make something that seems like it can be interpreted, but creates very strong conflicts with our prior knowledge that are almost impossible to overcome. To do this, I chose very direct choices of encoding, such as using color to represent color, and position to represent position, then set the mappings counterintuitively so I could wreak complete havoc with viewers’ minds.

我的目标是做出看起来似乎可以解释的内容，但与我们现有的知识产生非常强烈的冲突，而这几乎是无法克服的。 为此，我选择了非常直接的编码选择，例如使用颜色表示颜色，使用位置表示位置，然后反直觉地设置映射，这样我就可以完全破坏观看者的思想。

(Data)

I was looking for a good dataset to try the idea on but could not find one I really liked. Then I got the idea that it would be recursively bad to create a bad visualization, of bad visualizations, so I manually collected some data from viz.wtf

我一直在寻找一个很好的数据集来尝试这个想法，但是找不到我真正喜欢的数据集。 然后我想到了创建不良的可视化效果和递归的可视化效果将是递归的，所以我手动从viz.wtf收集了一些数据

(Here are all the crimes I have committed to this chart:)

  1. I used color to represent color, but didn’t guarantee that they would be the same color. As a result, green is the new black.
    我用颜色表示颜色，但不保证它们会是相同的颜色。 结果， 绿色就是新的黑色 。
  2. I also didn’t add enough unique colors, so there are duplicates. For instance, both red and blue are represented by green. (This was not intentional at first, but then it made things look worse so I kept it.)
    我也没有添加足够的独特颜色，因此存在重复项。 例如， 红色和蓝色都由绿色表示。 (起初这不是故意的，但随后使情况看起来更糟，因此我保留了它。)
  3. There was a special case for “mixed” color, as I couldn’t decide what color to encode it with. As a result, each of these “mixed” visualizations received a randomly selected color.
  4. I used position to represent position, but ensured that these never matched up. With that, right is on the left.
  5. I used shapes to represent chart types, but ensured that they never matched. With that, a bar chart is a circle, while a pie chart looks like a bar.
  6. I used size to encode popularity, but used an inverse scale with the biggest size meaning zero.
  7. I made axis labels more complicated than they needed to be. No 3D? True or false?
  8. I rotate each giant number by its value in degrees. This one is pointless encoding.
  9. The circles around the giant numbers don’t mean anything. They do not indicate boundaries.
  10. If you sum all the numbers, there are actually 102 visualizations, not 100.
  11. I added a dinosaur. Because I could.
  12. Lastly, there was a link to the raw data, proudly shared in PDF format.

翻译自: https://www.freecodecamp.org/news/how-i-carefully-crafted-a-terrible-visualization-2c8e06d50ebb/

ae制作数据可视化