地理信息的可视化

• 目的:

  • 认识数据
  • 展示数据

• 相关工具

  • 商务工具: Tableau(Excel)
  • GIS工具: ArcGIS
  • 前端工具: Echarts.js D3.js,...

相关的R包

• 静态地理信息:
  • maps
  • ggmap
  • choroplethr
• 交互地理信息:
  • leaflet
  • googleVis
• 辅助的R包:
  • RgoogleMaps, mapdata, mapproj, baidumap,...

静态化地图

• maps
• 最早的一批地理信息可视化
• US, world, china,...

library(maps)
library(mapdata)
data("chinaMapEnv")
map('china')

静态化地图

• choroplethr
• 海外很火的R包

library(choroplethr)
library(choroplethrMaps)
data(df_pop_country)
country_choropleth(df_pop_country,
                   title      = "2012 Population Estimates",
                   legend     = "Population",
                   num_colors = 1,
                   zoom       = c("china", "japan", 
                                  "north korea","south korea"))

实时静态地图

• ggmap

http://cos.name/2013/06/weibo-fans-map-visualization/

实时静态地图

• Sorry google http://nicpic.com

实时静态地图

• HI baidu
• 与ggplot2结合
• 数据获取

devtools::install.github("badbye/baidumap")

library(baidumap)
p <- getBaiduMap('南昌') # the same
ggmap(p)+ geom_point(aes(x=115.9,y=28.7),size = 3,color='red')

可交互地图

• leaflet
• recharts

library(leaflet)
m = leaflet() %>% addTiles()
m = m %>% setView(115.86453, 28.68767, zoom = 17)

REmap

• REmap是一个基于Echarts 的R包.

• 提供一个简便,可交互的地图数据可视化工具.目前托管在github

devtools::install_github('lchiffon/REmap')
library(REmap)

• 动态可交互地图
• 代码友好
• 解决常用的地图需求
• 期待更多的人接触Echarts

基本原理

• 使用R处理传入的数据
• 写出Javascript脚本(Echarts)
• 保存在临时的html文件中
• 调用浏览器展示

不需要的

• 不需要懂html
• 不需要懂Javascript/Echarts

需要的

• R
• Chorme或者Firefox作为默认浏览器
• UTF-8编码的数据框

分级统计图

• 使用颜色的深浅来展示数据
• 应用最广泛的地理位置可视化之一

数据来源: 中国统计年鉴2015年第二季度GDP

library(REmap)
GDP = read.csv("data/GDP.csv",encoding = "UTF-8")
remapC(GDP,
         color = c("red","yellow"),
         title = "2015年第二季度中国各省份GDP(亿元)",
         subtitle = "数据来源中国统计年鉴")

Link

数据来源: http://baike.baidu.com

popu = read.csv("data/jiangxi.csv",encoding = "UTF-8")
plotdata = data.frame(city = popu$city,
                             density = popu$population/popu$space*1e4)
remapC(data = plotdata,maptype = "江西",
         title = "江西人口密度(人/平方公里)")

Link

Bmap

remap(demoC,theme = get_theme("Bright"))

Link

remapB(markLineData = demoC,
         markPointData = demoC[,2])

# remapB(markLineData = demoC,
#        markPointData = demoC[,2],color = "Blue")

link

link

Bmap

城市级地图

## 获取线路
library(baidumap)
df = getRoute('南昌西站', '江西财经大学北区',region="南昌")
## 准备数据
df$label = paste("Point",1:dim(df)[1])
markLinedata = data.frame(origin = df$label[-dim(df)[1]],
                                  destination = df$label[-1])
## 绘制地图
remapB(center = get_city_coord("南昌"),
         zoom = 13,
         markLineData = markLinedata,geoData = df,
         markLineTheme = markLineControl(smoothness = 0,
                                                  effect = T,
                                                  symbolSize = c(0,0)))

link

## 获取数据
barbeque = getPlace('烧烤', '南昌')
## 准备数据
barbeque$tooltip = paste(barbeque$name,
                                 barbeque$address,
                                 barbeque$telephone,sep = "<br/>")
geodata = data.frame(barbeque$lon,
                            barbeque$lat,
                            barbeque$name)
## 绘制数据
remapB(center = get_city_coord("江西财经大学北区"),
         zoom = 15,
         markPointData = barbeque,
         geoData = geodata,
         markPointTheme = markPointControl(effect = F,symbol = "pin",
                                                     symbolSize = 7,
                                                     color="red"),
         title="南昌烧烤分布")

link

2015中国R语言会议

数据来源: http://cos.name

## 读取数据
rCon = read.csv("data/R-con.csv",encoding = "UTF-8")
## 准备数据
rCon$pic = paste0("<img src=",rCon$pic," width=300px />")
plotdata = data.frame(place = rCon$place,
                             symbol = "image://./pic/cos.png",
                             tooltip = paste(rCon$name,rCon$time,rCon$pic,sep = "<br/>"))
## 绘制地图
remapC(data = chinaIphone,
         color = c("white","white"),
         markPointData = plotdata,
         markPointTheme = markPointControl(effect = F,
                                                     symbolSize = 30),
         title = "2015中国R语言会议"
         )

link

最早的数据产品

什么算一个数据产品

• 数据驱动
• 建模预测
• 较好的可视化来展示结果

天气预报

• 数据驱动
• 建模预测
• 较好的可视化来展示结果
• 如果能更准确点...

cityvec = c("上海",
            "广州","大连","南宁","南昌","拉萨","长春","包头",
            "重庆","武汉","北京","洛阳","台北","乌鲁木齐","沈阳",
            "西安","成都","合肥","长沙","济南","福州","兰州")
weather = sapply(cityvec,get_weather_str)
plotdata = data.frame(cityvec,
                      tooltip=paste0(cityvec,weather[1,]),
                      symbol = weather[2,])
remapB(markPointData = plotdata,
       markPointTheme = markPointControl(symbolSize = 25,
                                         effect = F),
       color = "Blue",
       title = "未来天气预报",
       subtitle = "by REmap, Chiffon")

link

美好的未来

link

link