Recently a friend and work in government transportation departments chat, talk about big data applications, data, information technology in public transport management of their leadership attaches great importance in recent years, especially some of the large-screen monitor application traffic Kanban. Earlier in the face of demand for this type of data visualization to show large-screen monitor, usually carried out by the parties through codes outsourcing projects and other forms of development, and art through continuous optimization to adjust, and eventually delivered to the relevant departments for review and on-line.
In this way, although some very detailed style also needs to keep the code tuning, but the problem is caused large-screen billboards from requirements gathering to begin to develop, and then to the final on the line is too long a period, generally a big screen back and forth at least half a month time to complete the development, the current leadership of the lengthy development cycle is not satisfied.
As a done N items in BI engineers, after learning that a friend's problems, heart have been tempted to help impulsive. As more he needs big screen, waiting for the follow-up from the provincial to the municipal level has a number of projects approved, and the large amount of data. So I recommend: a small amount of custom code required visualization can be extremely cool professional high, and for the most updated often need a large screen can report / BI to develop.
Large screen is generally no difficulty data analysis, mainly visual appearance, performance for supporting large amount of data. But even these two requirements, developers must tired vomiting blood. BI to develop and use it, as long as the operating platform to build a good formation data, from the front end of data acquisition, data integration, making visualization are coherent.
For example, the following figure with FineBI to do a "big public transport data management platform", the data prepared under the premise of a skilled BI engineers probably spent more than an hour, he made out.
Specifically how to do, I have to be careful to say ~
About FineBI
FineBI is a self-service BI analysis tools to quickly connect data, cleaning up the data, make charts, do the visualization report. Mainly used for routine business reports (can quickly create), business data analysis. It is characterized by fast chips is simple, similar to the operation PivotTable, drag and drop data fields, filter screening nomogram can generate desired, compared to PivotTables Do not write VBA to write a formula.
Here we can first get through the official website FineBI software activation code, download and install the wizard:
I suggest that you refer to the official help documentation operations.
Production of large-screen steps
1. Data Modeling
First prepare the data. A plurality of data tables which generally requires the integration of the data relation table, get their required data fields.
和朋友那边拿到他们公共交通交通运输的相关数据（为了简单方便，直接导出为excel文件）之后，将要用到的公共交通路线表相关的excel数据表文件，上传导入到FineBI业务包中。
为了方便对维度表和事实表进行关联管理（1端一般是维度表，N端是事实表），FineBI可以对基础表灵活地建立对应关联关系，方便通过自助数据集对基础表的相关数据灵活地进行调用和加工建模。
另外一方面，通过可视化交互和操作自助数据集，能够实现以0编码的形式实现数据加工的过程（例如过滤、分组汇总、新增列、排序、左右合并等等），最终得到自己所需要的数据，具体大家可以看帮助文档摸索。
2.大屏数据可视化
数据准备好之后就要做图表设计可视化了。制作大屏需要一定的审美，先不说好看，起码要有最基本的布局意识（图表怎么分布，指标怎么展示）。
在给大家介绍具体制作过程之前先讲解一下通常大屏管理驾驶舱的布局方式。大屏数据往往展现的是一个企业全局的业务，一般分为主要指标和次要指标两个层次，主要指标反映核心业务，次要指标用于进一步阐述分析。所以在制作时给予不一样的侧重，这里推荐几种常见的版式。
上面几个版式不是金科定律，只是通常推荐的主次分布版式，能让信息一目了然。实际项目中，不一定使用主次分布，也可以使用平均分布，或者可以二者结合进行适当调整。比如下图所示，指标很多很多，存在多个层级的，就根据上面所说的基本原则进行一些微调，效果会很好。
这里我选择的是第4种样式，然后看看我是如何来做图表的。
- 图表可视化化操作：
在FineBI里，每一个图表都称之为图表组件，我们可对图表做单独的设计，比如配色，比如上传背景底图。每一个图表组件再配上一些文字组件，表格组件，组合起来就成为了dashboard，也就是我们要做的大屏。
How to make a chart (assembly), the line pressure to [maximum] top10 chart component as an example. We just need to route and number of passengers were dragged into horizontal and vertical areas, and then in descending order according to the number of passengers data prior to filtration 10, it can automatically generate a histogram line pressure top10 of the largest. Then the number of passengers field is bound to color the area, you can achieve the number of passengers descending line color gradient display.
Other chart is similar to the operation, the core is what you want to know every piece of the show. Here is not an overview, I do offer the various components of the chart:
- Layout:
1. First, we can add text component, add a title to the entire dashboard of a prominent, people will be able to see the whole theme to express the big screen data;
2. Since it is a public transport management screen, we can start with the core index is most concerned about the overall leadership of departure, for example, we can ride the bus the bus passengers as the large ones and the cumulative key indicators show, respectively, arranged symmetrically in the dashboard about location;
3. Departure time distribution and distribution destination, since it is a distribution statistics, we can prefer pie (For more beautiful, we set to ring chart) on display, you can see most of the departure time zone or distributed during the day, accounting for 85.25%;
4. distribution of number of trips one week, because the trend is against time, we chose a line graph on display, it can be seen weekly trip focused on the highest peak on Saturday;
The line pressure top10 maximum travel time of day distribution because it is the comparison shows that we are using histograms and bar charts were presented;
6. For each detail data bus lines, here we can directly use the form to unfold.
Finally, each component to the dashboard and drag, size adjustment drama, as follows:
The final layout adjustments, allowing other components are unified suspended above the flow map, the location of other components can refer to the figure at about fine-tuning, click the Preview template after trimming is completed, so that our public transport management data big-screen production is complete .
By FineBI such a large-screen co-manage data traffic, leadership and decision-making related businesses want to see the sales data at a glance, we have not really looking forward to want to try it?