在Shiny中计算日期范围的平均值和中位数

Question

想要计算仅按数据表的选定日期范围分组的数字变量的均值和中位数，而不是小叶数据。 传单地图有效（只需要缩小以查看假的长/拉图，但现在不担心）。

我为数据的中间数/平均值的总和创建了第二个数据帧df10 。

到目前为止，尝试改变输入函数为平均值创建单独的变量，但发现它很麻烦，不需要我的需要。

尝试使用colMeans(dataset()[,which(sapply(dataset(), class) != "Date")])这里Shiny计算数据框中列的平均值

错误是"invalid 'x' type in 'x && y" 。 它与colmeans有关

### Generate a dataset ###
start_date <- as.Date('2018-01-01')  
end_date <- as.Date('2019-05-10')   
set.seed(1984)
date1 <- as.Date(sample( as.numeric(start_date): as.numeric(end_date), 988, 
                         replace = T), origin = '1970-01-01')
group <- rep(letters[1:26], each = 38)
x1 <- runif(n = 988, min = 3.26, max = 10)
x2 <- runif(n = 988, min = 3.26, max = 10)
x3 <- runif(n = 988, min = 3.26, max = 10)
x4 <- runif(n = 988, min = 3.26, max = 10)
x5 <- runif(n = 988, min = 3.26, max = 10)
latitude <- runif(988,40.75042,50.75042)
longitude <- runif(988,-73.98928,-63.98928)

dataframe <- cbind(data.frame(date1,group,x1,x2,x3,x4,x5,latitude,longitude))

df10 <- cbind(data.frame(date1,group,x1,x2,x3,x4,x5))
library(lubridate)
dataframe$date <- ymd(dataframe$date1)
df10$date <- ymd(df10$date1)

library(shiny)
library(leaflet)
library(DT)
dataframe$defectrateLvl <- cut(dataframe$x1, 
                               c(3.26,6,100), include.lowest = T,
                               labels = c('3.26-6x','6x+')) 
beatCol <- colorFactor(palette = c('yellow', 'red'), dataframe$defectrateLvl)


ui <- fluidPage(
  dateInput(inputId = "date", label="Select a date", value = "2019-03-01", min = "2018-01-01", max = "2019-05-10",
            format = "yyyy-mm-dd", startview = "month",
            language = "en", width = NULL),
  leafletOutput("map"),
  fluidRow(
    dateRangeInput("daterange","Date range:",start=Sys.Date()-10, end=Sys.Date() -1),
    DT::dataTableOutput("tbl")
  )
)

server <- shinyServer(function (input, output,session) {
  dailyData <- reactive(dataframe[dataframe$date == format(input$date, '%Y/%m/%d'), ] )
  output$map <- renderLeaflet({
    dataframe <- dailyData()  # Added this in attempt to integrate
    dataframe %>% leaflet() %>% 
      setView(lng = -73.98928, lat = 40.75042, zoom = 10) %>%
      addProviderTiles("CartoDB.Positron", options = providerTileOptions(noWrap = TRUE)) %>%
      addCircleMarkers(
        lng=~dataframe$longitude, # Longitude coordinates
        lat=~dataframe$latitude, # Latitude coordinates
        #radius=~defectrateLvl, # Total count
        popup =~ dataframe$group,
        color = ~beatCol(dataframe$defectrateLvl),
        fillOpacity=0.5 # Circle Fill Opacity
      )
  })  
  output$tbl<-DT::renderDataTable({
    dataset <- reactive({df10 })
    dataset() %>% group_by(group) %>% 
      filter(date > input$daterange[1],
             date < input$daterange[2])
    #sapply(Filter(is.numeric, df6), mean)
    colMeans(dataset()[,which(sapply(dataset(), class) !="date","date1","group")])
  })

})


shinyApp(ui, server)

我希望数值变量可以通过均值进行汇总，如果可能的话，可以通过中位数进行汇总，但此时不太重要。 任何帮助将不胜感激。

Answer 1

该错误是由最后一个函数引起的。

colMeans(df[,which(sapply(df, class) !="date","date1","group")])

此代码将该函数应用于不属于类xy的所有列。 "date"或"group"是列名。

ColMeans还会生成一个数字向量，这会导致错误，因为DT只能显示矩阵或data.frame。 我为您提供了一个创建数据帧的代码。 但是在genrell中我会考虑使用dplyr来创建你的结果。 这更容易。

这是一个有效的解决方案，但是您必须更改dateinputs，因为预定义的选择会创建一个包含0行的data.frame。

library(lubridate)
library(shiny)
library(leaflet)
library(DT)
library(dplyr)

### Generate a dataset ###
start_date <- as.Date('2018-01-01')  
end_date <- as.Date('2019-05-10')   
set.seed(1984)
date1 <- as.Date(sample( as.numeric(start_date): as.numeric(end_date), 988, 
                         replace = T), origin = '1970-01-01')
group <- rep(letters[1:26], each = 38)
x1 <- runif(n = 988, min = 3.26, max = 10)
x2 <- runif(n = 988, min = 3.26, max = 10)
x3 <- runif(n = 988, min = 3.26, max = 10)
x4 <- runif(n = 988, min = 3.26, max = 10)
x5 <- runif(n = 988, min = 3.26, max = 10)
latitude <- runif(988,40.75042,50.75042)
longitude <- runif(988,-73.98928,-63.98928)

dataframe <- cbind(data.frame(date1,group,x1,x2,x3,x4,x5,latitude,longitude))

df10 <- cbind(data.frame(date1,group,x1,x2,x3,x4,x5))
dataframe$date <- ymd(dataframe$date1)
df10$date <- ymd(df10$date1)


dataframe$defectrateLvl <- cut(dataframe$x1, 
                               c(3.26,6,100), include.lowest = T,
                               labels = c('3.26-6x','6x+')) 
beatCol <- colorFactor(palette = c('yellow', 'red'), dataframe$defectrateLvl)


ui <- fluidPage(
    dateInput(inputId = "date", label="Select a date", value = "2019-03-01", min = "2018-01-01", max = "2019-05-10",
              format = "yyyy-mm-dd", startview = "month",
              language = "en", width = NULL),
    leafletOutput("map"),
    fluidRow(
        dateRangeInput("daterange","Date range:",start=Sys.Date()-10, end=Sys.Date() -1),
        DT::dataTableOutput("tbl")
    )
)

server <- shinyServer(function (input, output,session) {
    dailyData <- reactive(dataframe[dataframe$date == format(input$date, '%Y/%m/%d'), ] )
    output$map <- renderLeaflet({
        dataframe <- dailyData()  # Added this in attempt to integrate
        dataframe %>% leaflet() %>% 
            setView(lng = -73.98928, lat = 40.75042, zoom = 10) %>%
            addProviderTiles("CartoDB.Positron", options = providerTileOptions(noWrap = TRUE)) %>%
            addCircleMarkers(
                lng=~dataframe$longitude, # Longitude coordinates
                lat=~dataframe$latitude, # Latitude coordinates
                #radius=~defectrateLvl, # Total count
                popup =~ dataframe$group,
                color = ~beatCol(dataframe$defectrateLvl),
                fillOpacity=0.5 # Circle Fill Opacity
            )
    })  

    dataset <- reactive({df10 })

    output$tbl <-DT::renderDataTable({
        df <- dataset()

        df <- df %>% 
            group_by(group) %>% 
            filter(date > input$daterange[1],
                   date < input$daterange[2])
        #sapply(Filter(is.numeric, df6), mean)
        result <- data.frame(colMeans(df[which(sapply(df, class)=="numeric")]))
        colnames(result)[1] <- "Result"
        result
        #colMeans(df[,which(sapply(df, class) !="date","date1","group")])
    })

})


shinyApp(ui, server)