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

This article shows you how to analyze data visualizations. It is concise and easy to understand. It will definitely make your eyes shine. I hope you can learn something from the detailed introduction of this article.

Every second of our lives is generating data, as you are reading this article. What phone you used, model number, location, time you clicked on this article, how long it took to read the article... are all collected data.

When data of the same kind are collected together, even when other data are vertically connected, they can become big data. Depending on who uses the data and for what purpose, big data is not strictly categorized. But when you have huge amounts of data, how do you most directly react to the data situation? How to find data that needs further analysis as quickly as possible? After doing a series of analysis results, how to convince your readers or leaders easily and quickly?

We may be big data producers ourselves. They will also be users.

What is Data Visualization?

In fact, big data is a very empty concept, and the specific meaning has different meanings for different people. The basic process chain of big data includes: data mining, data cleaning, data storage, data analysis, and data display.

Data visualization is such a way, throughout the whole period of data analysis, in the early stage to help data analysts quickly present the overall data, find possible problem points, later through visualization, later data analysis results more quickly presented to the reader. Data visualization can interpret data using various combinations of graphics, icons, color variations, etc., allowing people to quickly understand complex relationships in one or more sets of data. Through data visualization, one can discern trends, patterns, discover specific problems, and even inform decisions.

Take the recent example of the new crown, you must be familiar with the following pictures:

Number of newly diagnosed new crowns nationwide (Photo source: Lilac Garden)

Distribution of the existing cumulative number of confirmed cases nationwide (Photo source: Lilac Garden)

Newly diagnosed data in China (Photo source: Lilac Garden)

The first and second graphs are examples of data visualization. The broken line chart and map intuitively show the daily number and distribution of new crown pneumonia in the third chart with images, while eliminating the reader's interpretation and processing of the specific numbers of the third icon. When the reader needs to know more about the specific numbers behind the figures, the reader can go to the third icon to study. And this is just the simplest example.

Why do we need data visualization?

From the above example, if big data can help us find patterns and trends, then data visualization is a way to visualize data. Data visualization is a step through data cleansing, data integration. Let's say you're a salesperson and you report sales results to the next level up. Your data may include customer name, customer address, purchase product, purchase model, purchase quantity, purchase date, purchase order number, delivery time, delivery method, sales amount, discount number, profit margin... What angle should you approach your data from? When you conclude that you need to strengthen the promotion of a certain product, do you think your leaders prefer to understand the results intuitively, or are they willing to spend 15 minutes reading your analysis page by page?

Changes in American public perceptions of the new crown virus are more specific and direct in the form of images than spreadsheets or words. (Source: 538)

So data visualization can simplify the human brain processing information, and draw conclusions as an effective means. Our brains are more likely to draw conclusions from visual representations such as images. Even if some analysts can draw conclusions from complex calculations and model designs, these conclusions are far less intuitive than those summarized directly in graphs.

Take the simplest example, the Beijing and Shanghai subway bus map can be called a data visualization. criss-crossing rail transit, different line stations, intersecting transfer points, if presented through text or tables, are far less intuitive than rail transit maps.

Shanghai Rail Transit Map (Photo Source: TravelChinaGuide)

Since orbital maps are all a kind of data visualization, you can let go of association. In fact, in many aspects of life, we are all involved in data visualization.

What is the difference between data visualization and data analysis?

Data visualization can easily be confused with data analysis, although there are similarities-both data visualization and data analysis present data in a visual interface.

When multiple visualizations are combined, they can show more information and even tell a complete time (Source: Center for Data Innovation)

yes, there are many differences between the two. Data analysis is an exploratory process. Because a lot of data analysts get the data, usually there is a specific question to discover, around this question, to carry out different tests, need enough patience to discover the use of a certain means, focus on analyzing some data, in order to reflect certain relationships, and answer the initial question. Data visualization is a part of data analysis. Data visualization can be used in the early and late stages to achieve more effective data analysis and more clear presentation of final analysis results. It can be broadly summarized as follows:

Use for different purposes. Data analysis can uncover certain underlying patterns, or trends, that can help predict certain future events. The early or late data used for data analysis can be used as a data source for data visualization. Data visualization can more intuitively present a local feature, more clearly show the impact of a variable, the early stage is to help data analysts understand the general data situation, find outliers. The later stage can be post-analysis data to better display the analysis results.

The relationship between the two is different. Data analysis is a combination of analysis and visualization to find certain conclusions. Sometimes, data analytics is the front end of data visualization, and data visualization presents the results of data analytics.

Different tools are used. Data analysis is generally done through prescriptive analytics and predictive analytics, diagnostic analytics. Tools used for data analysis include Excel, hive, Ploybase, SAP Business Intelligence, Presto, Trifacta, Clear Analytics, and more. Data visualization can be static or interactive, using tools such as Plotly, DataHero, Tableau, QlikView, ZingCHhart, etc.

Tableau's interactivity can be digitized. (Source: TABLEAU)

How can data visualization help with data analysis and for what purposes?

Data visualizations have many uses, and each type of data visualization can have different uses. Here are some of the most common cases of data visualization.

Time changes. The most basic and common method is to use temporal variation to show the variation of another variable. Note, however, that this does not mean that such visualizations have no value. This type of data visualization is common because most data has a temporal component. Therefore, the first step in many data analyses is to see how the data changes over time.

Determine the frequency. Determining frequency is also one of the fundamental uses of data visualization. Because it also applies to designing data involving time. If time is involved, in addition to examining how the data changes over time, look at whether the frequency of relevant events over time is logical in a unit of time.

Determine relationships (associations) between data. Identifying associations between data is one of the most valuable uses of data visualization. Understanding relationships in data is important, but without visualization it is difficult to determine the relationship between two variables.

Check the entire data. In market research, it is often possible to use data visualization to examine examples of data as a whole. Because marketers and salespeople need to know what demographic their message is aimed at, they need to analyze the audience for the entire market, as well as the associations between clusters within the demographic, influencers within the cluster, and outliers.

Time planning. When working on a very complex project or schedule, it usually involves different departments, different people, and different project details, which can be very confusing. A Gantt Chart can solve this problem by clearly stating each task in the project and the time it takes to complete it.

Analyze value and identify risk. Because there are so many correlations and so many different variables to consider in analyzing metrics like value and risk, it is difficult to accurately and effectively discern value and risk at a glance using a common spreadsheet of variables. Data visualization can be like color-coding formulas to show which opportunities are valuable and which are risky.

Four basic types of data visualization graphics (Source: The Coding Room) What are the types of data visualization?

There are many types of visualizations, and here I list most of them available on the market.

Line chart

Line chart (Source: New Zealand Census)

Area chart

Area map (Source: Wikipedia)

Bar chart

Bar chart (Source: Naomi Robbins)| Forbes)

Histogram

Histogram (Source: Naomi Robbins)| Forbes)

It is worth noting that there is a difference between bar charts and histograms. The width of the bar indicates category and is fixed, and the length indicates frequency. Histograms use "area" to represent the frequency of each group, height to represent the frequency of each group, and width to represent the group distance, so height and width are meaningful. The histogram is a continuous series of X axes, and is arranged continuously. The X axis of the bar chart is categorical data, arranged separately.

Scatterplot

Scatter plot (Source: Wikimedia)

Box Plot

Box diagram (Source: Wikimedia)

bubble chart

Bubble chart (Source: Tony Hirst)| Flickr)

Pie chart

pie chart (Source: Wikimedia)

gauge

Gauge Map (Source Ken Flerlage)| The FlerlageTwins)

Maps

map

Center for Geographic Analysis - Harvard University

https://gis.harvard.edu/researchhttp://worldmap.harvard.edu/africamap/

and Harvard University's map of Africa:

http://worldmap.harvard.edu/africamap/ This interactive map includes economic, religious, social, demographic, historical, transportation and more.

Heat map

There are many kinds. Here's a heat map from eye tracking that I used.

Heat map function of eye tracker (Source: Rosenfeld Media)| Flickr

Frame diagram

A frame diagram is usually a tree diagram (Source: Wikimedia)

Waterfall chart

Waterfall map (Source: Wikipedia)

Funnel chart

Funnel graph created using R (Source: Neha Kuma)| Sisense)

Radar or Spider Chart

Radar map (Source: middlebury.edu)

These are the data visualizations that you can see on the market. Of course, there are other forms, as well as the superposition of graphics and graphics, such as the combination of line charts and histograms, etc., I will not list them one by one.

Therefore, data visualization is not so mysterious, we commonly use EXCEL can make several of the above visual graphics. Tableau can create most of these data visualizations in addition to Excel, and interactive data visualizations can be created by using Dashboard, and story functionality enables graph groups to create storytelling. The most important thing is that Tableau is available in a free version and uses everything Tableau has to offer. The difference between the free version and the paid version is that you can save it locally, while the free version can upload to Tableau's public resources, share your data visualizations with others, and view all the great visualizations that others have created.

That's how to analyze data visualizations. Have you learned anything or skills? If you want to learn more skills or enrich your knowledge reserves, please pay attention to the industry information channel.

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.