對(duì)數(shù)據(jù)可視化的理解

Data is weaving its way into almost all aspects of our lives since the past decade. Our ability to store more information in smaller and smaller spaces has encouraged us to make sure we leave no information out. The ease of collecting information, and the need to predict future events based on past behavior has allowed integration of data collection in all walks of life. From the number of views on a web page to the number of steps we take, we want to know every single metric that can give us better insight into the lives we lead.

自從過去十年以來，數(shù)據(jù)已經(jīng)滲透到我們生活的幾乎所有方面。 我們?cè)谠絹碓叫〉目臻g中存儲(chǔ)更多信息的能力鼓勵(lì)了我們確保不遺漏任何信息。 信息收集的便捷性以及基于過去行為來預(yù)測(cè)未來事件的需求已使各行各業(yè)的數(shù)據(jù)收集得以整合。 從網(wǎng)頁上的視圖數(shù)量到我們采取的步驟數(shù)量，我們希望了解可以幫助我們更好地洞察自己生活的每個(gè)指標(biāo)。

While the volume of data collected is increasing, our ability to extract value from this data is not able to match pace. It is not possible to glance at and process millions of rows of data — even when these rows are aggregated to a few hundred odd records.

隨著收集的數(shù)據(jù)量不斷增加，我們從這些數(shù)據(jù)中提取價(jià)值的能力無法與時(shí)俱進(jìn)。 不可能瀏覽和處理數(shù)百萬行的數(shù)據(jù)，即使這些行匯總到幾百個(gè)奇數(shù)記錄中也是如此。

This calls for the need of data visualization — a method of graphically representing information that plays on the inbuilt intuitions and perceptions of humans to communicate large amounts of information in a relatively short period of time.

這就需要數(shù)據(jù)可視化-一種以圖形方式表示信息的方法，該信息以人類固有的直覺和感知為基礎(chǔ)，以便在相對(duì)較短的時(shí)間內(nèi)傳達(dá)大量信息。

Data specialists use programs and tools to comb through stores of data and detect valuable patterns from them. Visualizing outcomes using graphs and charts eases the cognitive load on our brains due to their pre-attentive attributes — visual properties that our brains perceive without any conscious effort to do so. These attributes are divided into 3 major categories that aid our processing of information — color, form, and movement.

數(shù)據(jù)專家使用程序和工具來梳理數(shù)據(jù)存儲(chǔ)并從中檢測(cè)出有價(jià)值的模式。 使用圖形和圖表可視化結(jié)果可以減輕我們大腦的認(rèn)知負(fù)擔(dān)，這是因?yàn)樗鼈兙哂蓄A(yù)先注意的屬性-我們的大腦無需任何有意識(shí)的努力即可感知的視覺特性。 這些屬性分為3個(gè)主要類別，可幫助我們處理信息- 顏色，形式和運(yùn)動(dòng)。

Let’s delve a little deeper into understanding what these mean with an example.

讓我們通過一個(gè)示例來更深入地了解這些含義。

The data set used is the Superstore Data — a dummy list of orders from a store based in the United States. The data set consists of the sales and profits of items from different categories across the country, which we can represent visually through a tool like Tableau.

所使用的數(shù)據(jù)集是“ 超級(jí)商店數(shù)據(jù)” -來自美國(guó)一家商店的訂單的虛擬列表。 數(shù)據(jù)集由全國(guó)不同類別的商品的銷售和利潤(rùn)組成，我們可以通過Tableau之類的工具直觀地表示這些商品。

顏色 (Color)

Different colors and hues help us separate information elements from their surroundings and from each other. This saves us the effort of searching through data serially.

不同的顏色和色調(diào)可幫助我們將信息元素與周圍環(huán)境以及彼此分開。 這節(jié)省了我們串行搜索數(shù)據(jù)的精力。

Colors in gradation are also helpful in conveying intensity. Representing a single metric (profit per state) on a map can communicate states with low and high profit based on the shade of green it is filled with. Our brains automatically infer that a lighter shade indicates lower profit while a darker one indicates higher profit.

漸變色也有助于傳達(dá)強(qiáng)度。 在地圖上表示單個(gè)度量標(biāo)準(zhǔn)(每個(gè)州的利潤(rùn))可以根據(jù)填充的綠色陰影傳達(dá)高低利潤(rùn)的州。 我們的大腦會(huì)自動(dòng)推斷出，較淺的陰影表示較低的利潤(rùn)，而較深的陰影則表示較高的利潤(rùn)。

Colors also have cultural connotations associated with them. A dark red color universally represents danger but when used in combination with green, it can mean “No” (to green’s “Yes”).

顏色還具有與之相關(guān)的文化內(nèi)涵。 深紅色通常表示危險(xiǎn)，但是與綠色結(jié)合使用時(shí)，它的意思是“否”(綠色表示“是”)。

形成 (Form)

The form of visual elements can be used to draw attention towards or away from it. Size, shape, grouping, length are all various types of forms that our brains grasp without any effort.

視覺元素的形式可以用來吸引人們注意或遠(yuǎn)離它。 大小，形狀，分組，長(zhǎng)度是我們的大腦無需費(fèi)力即可掌握的各種類型的形式。

In the map above, notice how we can combine both color and form cues together to highlight both Sales and Profit in different states.

在上面的地圖中，請(qǐng)注意我們?nèi)绾螌㈩伾托问教崾窘Y(jié)合在一起以突出顯示不同狀態(tài)的“銷售”和“利潤(rùn)”。

However, it is important to avoid overuse of too many cues at the same time, because it can overwhelm the viewer with details (which defeats the purpose of effective visualizations).

但是，重要的是要避免同時(shí)使用過多提示，因?yàn)樗赡苁共榭凑卟恢?這違背了有效可視化的目的)。

Grouping is another form that relies on our ability to chunk separate markers together. It is based on the Gestalt Principles of Psychology describing how humans typically see objects by grouping similar elements, recognizing patterns and simplifying complex images.

分組是另一種形式，它依賴于我們將單獨(dú)的標(biāo)記塊在一起的能力。 它基于格式塔心理學(xué)原理 ，描述了人們通常如何通過對(duì)相似元素進(jìn)行分組，識(shí)別模式并簡(jiǎn)化復(fù)雜圖像來看到物體。

The scatter plot above highlights the relationship between Sales and Profit of different categories in the Superstore. We automatically group the dots based on their color and proximity. This can help us come up with insights like — The furniture category (in blue) has high sales yet low profit margins.

上面的散點(diǎn)圖突出顯示了超市中不同類別的銷售和利潤(rùn)之間的關(guān)系。 我們會(huì)根據(jù)點(diǎn)的顏色和接近程度自動(dòng)將其分組。 這可以幫助我們得出一些見解，例如-家具類別(藍(lán)色)具有很高的銷售額但利潤(rùn)率很低。

運(yùn)動(dòng) (Movement)

Flashing and motion are two types of movements that draw our attention. The presence of motion helps us understand progress from one point to another. Flashing, or flickering, is used to call attention to a particular element immediately. It is advisable to use flashing only when absolutely necessary, since it can become annoying to have our eyes drawn to a certain place when we want to focus somewhere else.

閃爍和動(dòng)作是引起我們注意的兩種動(dòng)作。 運(yùn)動(dòng)的存在有助于我們理解從一個(gè)點(diǎn)到另一個(gè)點(diǎn)的進(jìn)展。 閃爍可用于立即引起人們對(duì)特定元素的注意。 建議僅在絕對(duì)必要時(shí)才使用閃光燈，因?yàn)楫?dāng)我們要聚焦到其他地方時(shí)，將眼睛吸引到某個(gè)特定位置會(huì)變得很煩人。

結(jié)論 (Conclusion)

The science behind our thoughts, learning and mental structures is extensively researched under the discipline of Cognitive Science and modern visualization tools like Power BI and Tableau help in applying years of research to enrich visual processing of data effectively.

我們的思想，學(xué)習(xí)和心理結(jié)構(gòu)背后的科學(xué)在認(rèn)知科學(xué)的學(xué)科下得到了廣泛的研究，諸如Power BI和Tableau之類的現(xiàn)代可視化工具有助于應(yīng)用多年的研究來有效地豐富數(shù)據(jù)的視覺處理。

…………………………………………………………………………………

………………………………………………………………………………

I am an independent Data Analyst competent in Tableau, Python and SQL actively soliciting contract, freelance or remote opportunities. I can be reached at unnati.palan@gmail.com (Linkedin | Tableau Public | Github)

我是Tableau，Python和SQL方面的獨(dú)立數(shù)據(jù)分析師，擅長(zhǎng)積極尋求合同，自由職業(yè)者或遠(yuǎn)程機(jī)會(huì)。 可以通過unnati.palan@gmail.com與我聯(lián)系( Linkedin | Tableau Public | Github )

翻譯自: https://medium.com/analytics-vidhya/make-data-visualizations-easier-to-understand-17422be3f931

對(duì)數(shù)據(jù)可視化的理解

總結(jié)

以上是生活随笔為你收集整理的对数据可视化的理解_使数据可视化更容易理解的全部?jī)?nèi)容，希望文章能夠幫你解決所遇到的問題。

如果覺得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。