Shulou(Shulou.com)06/01 Report--

This article mainly introduces "which excellent visualization and visualization projects managed by big data". In daily operation, I believe many people have doubts about which excellent visualization and visualization projects managed by big data. The editor consulted all kinds of materials and sorted out simple and easy-to-use operation methods. I hope it will be helpful for you to answer the doubts about the excellent visualization and visualization projects managed by big data. Next, please follow the editor to study!

There are many "rules" for visualization. Some are practical rules, some are suggestions to help you make a choice. If it's because of data requirements, and you know what to do, then many practical rules don't have to be followed.

However, there are some rules that should not be broken. These rules are usually used for certain kinds of charts that can only be read in a specific way. When these rules are broken, the data may be misread during the reading process. This is gonna be a little tricky.

The baseline of the bar chart must start from zero

Bar charts rely on length to render data. Short bars represent lower values, while longer bars represent higher values. The principle of a bar chart is to compare the size of the value by comparing the length of the bar.

When the baseline is changed, the visual effect is distorted.

For example, look at the picture above. The first bar chart on the left compares two values: 50 and 100, it has one and it has a baseline starting with zero. Fine. The length of the bar representing the number 100 is exactly twice the length of the number 50, and 100 is exactly twice the size of 50.

But when you change the baseline to a higher, non-zero value, the length of the first bar becomes shorter, while the length of the other bar does not change. A bar with a value of 100 is no longer twice as long as a bar with a value of 50. And so on, when the bar on the left representing the number 50 disappears completely, it means that 100 is infinitely greater than 50.

The baseline of the bar chart must start from zero.

Ex.: this bar chart is approved by Fox News.

On March 31st, the target value was 77.066 million, 17.8% higher than 6000000, but the second bar was almost three times the length of the first bar.

One might retort that the point of this picture is the difference between the two values rather than the two values themselves. Even so, it is a wrong choice to represent it with a bar chart. It might be better to use time series to present monthly accumulations.

Don't be too enthusiastic about pie charts.

Some people think that pie charts should be avoided altogether. Maybe they're right, maybe they're not. Some people may say that using a pie chart is an unforgivable mistake. I do not agree with this. Anyway, the fact is that people still use pie charts, so we can at least try to use them correctly.

Avoid over-cutting the pie chart, or it will be difficult to read it eventually.

So how much is "too much"? It's a matter of judgment. However, if it is already difficult to see from the picture that one of the fans is twice the size of the other, or several smaller fan-shaped areas look about the same size, it is time to stop on the fan-shaped cutting. At this point, consider grouping smaller categories into a larger one: "other". The same is true of the circle chart.

At the same time, consider using other kinds of charts to represent the scale.

Don't rely too much on pie charts.

Ex.: this pie chart is from Wikipedia. It shows different regions of the country.

Many pieces have been cut in the pie chart on the left, but another pie chart is separated next to it, showing the situation of smaller countries that cannot be seen clearly on the left, so as to provide more information. There are many ways to display this set of data, such as tree views, icons based on data scale, or just plain maps. A thin pie chart is only suitable for displaying data with only a few sets of values.

Respect the proportion of the part as a whole

Compared with the presented values, some graphs focus more on the relationship between the part and the whole, and they show the proportion of the part as a whole. For example, stacked bars, stacked areas, trees, mosaics, donuts, and pie charts. In these charts, each part represents a separate, non-overlapping scale.

With regard to this article, the most common error occurs when multiple choices are allowed in the survey question. For example: "what kind of transportation did you use last week? you can choose more." In this way, there will be a proportional overlap on the issue of people's multiple choices, and the sum of the percentages of different options is greater than one. To avoid this, you can't just make a statistical chart of the scale.

Ex.: this pie chart is from Fox News, which shows three percentages that do not belong to the same whole.

Each value is a separate whole, so in this case, the proportion of each value is more intuitively represented with three stacked bars (or ordinary bars).

Display data

Let the reader see the data, which is the focus of visualization. If the presentation of the data is not clear enough, it goes against the original intention of making a chart. This is often because there is too much data in a picture, so the reader's interest is distracted.

This is a classic "overdrawing" problem, and there are many related studies. But for basic diagrams, there are also some simple solutions.

First of all, you can change the size of the symbol so that the dots (or other symbols) in the image above don't take up too much space. In order to make the data intuitive and clear, it is necessary to increase the blank space.

Adjust the transparency so that the multi-level pattern will not be covered.

Divide the population into smaller subgroups by sampling or classifying the data. From this, you can take a smaller and more approach, so that there will be less information in each table.

The data are re-counted and classified.

All in all, better present the data.

Ex.: this picture shows every shot by the Golden State Warriors in 2008-09.

This picture eventually forms the shape of a court and draws a small conclusion about where the players shoot the most-the near frame, the middle distance, and the three-pointers. But the gap between them is so small that readers can't see the real difference of magnitude.

Data aggregation will help to solve such problems.

Interpretive coding

The data is presented through the combination of certain shapes, colors and geometry. In order for readers to read clearly, chart designers have to decode these graphics back to data values. A classic example is an unlabeled axis.

Sometimes the coding doesn't need to be explained. For example, the reader may know how to read a bar chart and do not have to explain that the length of the bar represents the size of the value. But the designer should really explain the data, that is, the units and themes of the chart.

So indicate what the axes represent. To provide the reader with clues or legends, explain the chart.

Ex.: this incorrectly marked chart is from the Winnipeg Sun:

If only we could know what the statistics are about.

Top Ten projects for data Visualization in 2015

Excellent visualization works will be full of life in 2015, and I'm sure there will be a lot of good works next year. A large number of projects have sprung up across different themes and applications, but if I were asked to choose an annual theme, it must be "teaching", whether through explanation, simulation or in-depth analysis. Sometimes it feels that visual creators are bold and try to get readers to stop thinking about data and statistics in the usual way. I like that very much.

Here are my best projects of 2015. According to convention, it is ranked in no particular order. At the same time, there are a lot of works that are not on this list, and they are also excellent.

Let's take a look at them.

1. Dear data.

This is an interesting project worth tracking, and its two topics-visualization and self-monitoring-have caught my attention.

Dear data is an one-year project done by Stefanie Posavec and Giorgia Lupi. Everyone tracks what happens every day of the week, such as how many times each person answers the phone, and then visualizes the data on a postcard. Then they send the postcards to each other-Lupi now lives in New York and Posavec lives in London.

two。 You draw: how does family income predict a child's chances of going to college?

It feels like a year to challenge readers how to understand data from a higher statistical perspective.

Gregor Aisch, Amanda Cox and Kevin Quealy of the New York Times Upshot asked readers to draw a line reflecting the percentage of household income to that of college children. So you can see how the lines you draw, the lines drawn by the real data, and what other people think of the relationship.

For Upshot/New York Times and data assumptions, see this quick puzzle to test your pattern-finding ability (quick test of your pattern recognition ability) and 3murd chart of the economy's future (3D map of economic future development).

3. "Black" brings out your scientific glory

New York Times 538 also carries out some data science teaching activities through their visual interaction. One of the best lessons is how to steal P values (p-hacking) to get the results you want from the same dataset.

The project was launched at a time when a recent graduate was exposed to falsify data (hyperlink). The focus of Christie Aschwanden and Ritchie King is not to doubt how an absurd result passed rigorous peer review. Instead, what they want to say is that doing research scientifically and interpreting data is the real difficulty.

4. Make the song "where are you now?"

The team from the New York Times did an interview about Justin Bieber, which was not only good, but also interesting.

Although Bieber is more like a supporting role, because Diplo and Skrillex are the ones who study how to make a hit song in detail, this music visualization next to the video can help you better understand what musicians are talking about.

5. How and when measles spreads among vaccinated children

The Guardian's Rich Harris,Nadja Popovich and Kenton Powell show what happens when children in a country are not vaccinated against measles.

As a parent, I want to make this whole list interactive.

6. Visual introduction of Machine Learning

Machine learning seems like a magical concept, as if it means that a robot can do strange things without your instruction. Stephanie Yee and Tony Chu use a visual example to solve the mystery.

This visual example takes you step by step through how the machine "learns" like a scroll. The transition chart makes the whole picture very smooth. The current results seem to be the first part of a series of projects, but we may have to wait a while to see the rest.

7. The end of the second World War

Part of Neil Halloran's project is recording, the other part is interactive visualization, the two seamlessly link together.

I was surprised to find that not many people do this type of project. When I realize that they are working on such a project, I very much hope that this will continue (read more).

8. 2014 was the hottest year on record

The most intuitive way to visualize this kind of data is a single line diagram. But after decomposing the line, we can get more information.

This dynamic chart by Tom Randall and Blacki Migliozzi of Bloomberg shows the monthly average temperature. Each line represents a complete year, and the line rises by a few inches over time from far to near.

9. Network effect

It's been a while since the last time I saw a project like Jonathan Harris.

He worked with Greg Hochmuth on the project, "Network effect" is a review of the entire Internet, allowing people to learn about all aspects of the Internet in a wonderful and fascinating way, for a few minutes at a time.

10. Commonly used metaphors

Figurative rhetoric is commonly used by writers and has a specific image, a tool and a writing habit in the reader's mind. The "Common metaphors" project done by the Bocoup data Visualization team studies the words commonly used in figurative rhetoric.

If you want to understand the gender roles and characters in the movie, this project is exactly what you are looking for.

At this point, the study of "which websites big data manages excellent visualization and visualization projects" is over. I hope to be able to solve your doubts. The collocation of theory and practice can better help you learn, go and try it! If you want to continue to learn more related knowledge, please continue to follow the website, the editor will continue to work hard to bring you more practical articles!

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.