added dataset + shiny app with map, barplot and data table
This commit is contained in:
patrycjalazna
parent
2350e8fdd5
commit
cd5275b075
4976
proj3/airplane_crashes.csv
Executable file
4976
proj3/airplane_crashes.csv
Executable file
File diff suppressed because it is too large
Load Diff
240
proj3/projekt.R
Normal file
240
proj3/projekt.R
Normal file
@ -0,0 +1,240 @@
|
|||||||
|
# setup srodowiska ####
|
||||||
|
setwd('/Users/patrycjalazna/Desktop/wizualizacja-danych/projekt_3/')
|
||||||
|
install.packages('shinyjs')
|
||||||
|
|
||||||
|
# importy ####
|
||||||
|
library(shiny)
|
||||||
|
library(shinyTime)
|
||||||
|
library(shinyjs)
|
||||||
|
library(highcharter)
|
||||||
|
library(splusTimeDate)
|
||||||
|
library(tidyverse)
|
||||||
|
library(lubridate)
|
||||||
|
library(gridExtra)
|
||||||
|
library(usmap)
|
||||||
|
library(tidytext)
|
||||||
|
library(tm)
|
||||||
|
library(SnowballC)
|
||||||
|
library(wordcloud)
|
||||||
|
library(dplyr)
|
||||||
|
library(streamgraph)
|
||||||
|
library(plotly)
|
||||||
|
library(DT)
|
||||||
|
|
||||||
|
|
||||||
|
# zaladowanie datasetu ####
|
||||||
|
airplane_crashes <- read.csv('airplane_crashes.csv')
|
||||||
|
|
||||||
|
# podsumowanie ####
|
||||||
|
summary <- summary(airplane_crashes)
|
||||||
|
colnames(airplane_crashes)
|
||||||
|
|
||||||
|
# preprocessing ####
|
||||||
|
# zmiana nazwy kolumn
|
||||||
|
colnames(airplane_crashes) <- c('Date', 'Time', 'Location', 'Operator', 'Flight', 'Route', 'Type',
|
||||||
|
'Registration', 'cn/ln', 'Total_Onboard', 'Passengers_Onboard',
|
||||||
|
'Crew_Onboard', 'Total_Fatalities', 'Passengers_Fatalities',
|
||||||
|
'Crew_Fatalities', 'Ground', 'Summary')
|
||||||
|
|
||||||
|
# konwersja na prawidlowy format Date
|
||||||
|
airplane_crashes$Date <- mdy(airplane_crashes$Date)
|
||||||
|
airplane_crashes$Time <- hm(airplane_crashes$Time)
|
||||||
|
|
||||||
|
# ekstrakcja informacji z datasetu, zapis w poprawnym formacie
|
||||||
|
# ekstrakcja informacji z kolumny Date
|
||||||
|
airplane_crashes$Year <- year(airplane_crashes$Date)
|
||||||
|
airplane_crashes$Month <- as.factor(month(airplane_crashes$Date))
|
||||||
|
airplane_crashes$Day <- as.factor(day(airplane_crashes$Date))
|
||||||
|
airplane_crashes$Weekday <- as.factor(wday(airplane_crashes$Date))
|
||||||
|
airplane_crashes$Is_Leap_Year <- leap_year(airplane_crashes$Date)
|
||||||
|
airplane_crashes$Decade <- year(floor_date(airplane_crashes$Date, years(10)))
|
||||||
|
airplane_crashes$Date <- format(as.Date(airplane_crashes$Date, '%m/%d/%Y'), '%d/%m/%Y')
|
||||||
|
|
||||||
|
# ekstrakcja godziny, minuty i AM/PM z kolumny Time
|
||||||
|
airplane_crashes$Hour <- as.integer(hour(airplane_crashes$Time))
|
||||||
|
airplane_crashes$Minute <- as.factor(minute(airplane_crashes$Time))
|
||||||
|
airplane_crashes$AM_PM <- if_else(am(airplane_crashes$Time), 'AM', 'PM')
|
||||||
|
airplane_crashes$btwn_6PM_6AM <- if_else
|
||||||
|
|
||||||
|
|
||||||
|
# ShinyApp ####
|
||||||
|
load_data <- function() {
|
||||||
|
Sys.sleep(2)
|
||||||
|
hide("loading_page")
|
||||||
|
show("main_content")
|
||||||
|
}
|
||||||
|
|
||||||
|
ui <- fluidPage(
|
||||||
|
useShinyjs(),
|
||||||
|
div(
|
||||||
|
id = "loading_page",
|
||||||
|
h1("Loading...")
|
||||||
|
),
|
||||||
|
hidden(
|
||||||
|
div(
|
||||||
|
id = "main_content",
|
||||||
|
navbarPage("Airplane crashes from 1908 to 2020",
|
||||||
|
tabPanel("General overview",
|
||||||
|
sidebarLayout(
|
||||||
|
sidebarPanel(
|
||||||
|
# wybór daty
|
||||||
|
dateRangeInput('dates',
|
||||||
|
'Date range:',
|
||||||
|
min(airplane_crashes$Date),
|
||||||
|
max(airplane_crashes$Date)),
|
||||||
|
# wybor godziny
|
||||||
|
timeInput("time", "Time (local):", seconds = FALSE)
|
||||||
|
),
|
||||||
|
mainPanel(
|
||||||
|
# wykresiki
|
||||||
|
plotlyOutput("weekdayCrashes")
|
||||||
|
|
||||||
|
)
|
||||||
|
)
|
||||||
|
),
|
||||||
|
tabPanel("Map",
|
||||||
|
mainPanel(
|
||||||
|
# mapy
|
||||||
|
plotlyOutput("mapPlot")
|
||||||
|
#plotOutput("mapPlot")
|
||||||
|
)
|
||||||
|
),
|
||||||
|
tabPanel("Data",
|
||||||
|
sidebarPanel(
|
||||||
|
dateRangeInput('dataDates',
|
||||||
|
'Date range:',
|
||||||
|
min(airplane_crashes$Date),
|
||||||
|
max(airplane_crashes$Date))
|
||||||
|
),
|
||||||
|
mainPanel(
|
||||||
|
DT::dataTableOutput('allData')
|
||||||
|
))
|
||||||
|
)
|
||||||
|
)
|
||||||
|
)
|
||||||
|
)
|
||||||
|
|
||||||
|
server <- function(input, output, session) {
|
||||||
|
load_data()
|
||||||
|
# godzina
|
||||||
|
observe(print(strftime(input$time, "%R")))
|
||||||
|
|
||||||
|
output$weekdayCrashes <- renderPlotly({
|
||||||
|
week_day <- airplane_crashes %>% count(Weekday)
|
||||||
|
|
||||||
|
week_day$Weekday[week_day$Weekday == "1"] <- "Monday"
|
||||||
|
|
||||||
|
|
||||||
|
# c('Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday', 'Sunday')
|
||||||
|
plot_ly(
|
||||||
|
x = week_day$Weekday, y = week_day$n,
|
||||||
|
name = 'Week Day Crashes', type = "bar"
|
||||||
|
) %>%
|
||||||
|
layout(title = 'Week Day Crashes',
|
||||||
|
xaxis = list(title = "Weekdays",
|
||||||
|
zeroline = FALSE),
|
||||||
|
yaxis = list(title = "No. of crashes",
|
||||||
|
zeroline = FALSE))
|
||||||
|
|
||||||
|
})
|
||||||
|
|
||||||
|
output$mapPlot <- renderPlotly({
|
||||||
|
|
||||||
|
states_list <- c('Alabama','Alaska','Alaksa','Arizona','Arkansas',"California",
|
||||||
|
"Colorado", "Connecticut","Delaware","Florida","Georgia",
|
||||||
|
"Hawaii","Idaho","Illinois", "Indiana","Iowa","Kansas",
|
||||||
|
"Kentucky","Louisiana","Maine","Maryland", "Massachusetts",
|
||||||
|
"Massachusett", "Michigan","Minnesota","Mississippi","Missouri",
|
||||||
|
"Montana", "Nebraska","Nevada","New Hampshire","New Jersey",
|
||||||
|
"New Mexico","New York", "North Carolina","North Dakota","Ohio",
|
||||||
|
"Oklahoma", "Oklohoma", "Oregon","Pennsylvania", "Rhode Island",
|
||||||
|
"South Carolina",
|
||||||
|
"South Dakota",'Tennesee',"Tennessee","Texas","Utah", "Vermont",
|
||||||
|
'Virginia',"Washington D.C.", "Washington, D.C.", "Washington",
|
||||||
|
"West Virginia","Wisconsin","Wyoming",
|
||||||
|
"AL", "AK", "AZ", "AR", "CA", "CO", "CT", "DC", "DE", "FL", "GA",
|
||||||
|
"HI", "ID", "IL", "IN", "IA", "KS", "KY", "LA", "ME", "MD",
|
||||||
|
"MA", "MI", "MN", "MS", "MO", "MT", "NE", "NV", "NH", "NJ",
|
||||||
|
"NM", "NY", "NC", "ND", "OH", "OK", "OR", "PA", "RI", "SC",
|
||||||
|
"SD", "TN", "TX", "UT", "VT", "VA", "WA", "WV", "WI", "WY")
|
||||||
|
|
||||||
|
location <- airplane_crashes %>%
|
||||||
|
select(Location)
|
||||||
|
|
||||||
|
for(state in states_list) {
|
||||||
|
location <- location %>%
|
||||||
|
mutate(Location = str_replace_all(Location, state, paste(state, ', USA', sep = ''))) %>%
|
||||||
|
mutate(Location = str_replace_all(Location, 'USA.*, ', '')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'West Virginia, USA,', '')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'Afghanstan', 'Afghanistan')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'Airzona|Arazona', 'Arizona')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'Alakska', 'Alaska')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'Cailifornia|Calilfornia', 'California')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'D.*Congo', 'DR Congo')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'Domincan Republic', 'Dominican Republic')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'Hati', 'Haiti')) %>%
|
||||||
|
mutate(Location = str_replace(Location, ' International Airport', '')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'Morrocco|Morroco', 'Morocco')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'Phillipines', 'Phillipines')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'Burma', 'Myanmar')) %>%
|
||||||
|
mutate(Location = str_replace(Location, '([Ss]outhern|[Nn]orthern|[Ww]estern|[Ee]astern) ', ''))}
|
||||||
|
|
||||||
|
country_state <- location %>%
|
||||||
|
select(Location) %>%
|
||||||
|
filter(!str_detect(Location, '[Oo]cean|[Ss]ea|[Cc]hannel|Gulf of')) %>%
|
||||||
|
mutate(Location = str_replace(Location, '(Near|Off|Over) ', '')) %>%
|
||||||
|
mutate(Location = str_replace(Location, 'USA, Australia', 'Australia')) %>%
|
||||||
|
mutate(State_Province = str_replace(Location, '(.*, )?(.*), (.*)', '\\2')) %>%
|
||||||
|
mutate(Country = str_replace(Location, '.*,\\s*', ''))
|
||||||
|
|
||||||
|
cntry <- country_state %>%
|
||||||
|
group_by(Country) %>%
|
||||||
|
summarize(n = n()) %>%
|
||||||
|
arrange(desc(n))
|
||||||
|
|
||||||
|
# podzial na kategorie wg liczby wypadkow w danym kraju
|
||||||
|
cntry <- cntry %>%
|
||||||
|
mutate(m = case_when(
|
||||||
|
n > 200 ~ "200 +",
|
||||||
|
n < 200 & n >= 100 ~ "199 - 100",
|
||||||
|
n < 100 & n >= 50 ~ "99 - 50",
|
||||||
|
n < 50 & n >= 10 ~ "49 - 10",
|
||||||
|
n < 10 ~ "< 10")) %>%
|
||||||
|
mutate(m = factor(m, levels = c("< 10", "49 - 10", "99 - 50", "199 - 100", "200 +")))
|
||||||
|
|
||||||
|
|
||||||
|
world_map <- map_data("world")
|
||||||
|
|
||||||
|
map_data <- cntry %>%
|
||||||
|
full_join(world_map, by = c('Country' = 'region'))
|
||||||
|
|
||||||
|
map_pal = c("#E0DFD5", "#E4B363", "#E97F02", '#EF6461', '#313638')
|
||||||
|
|
||||||
|
ggplot_obj <-
|
||||||
|
ggplot(map_data, aes(x = long, y = lat, group = group, fill = m)) +
|
||||||
|
geom_polygon(colour = "white") +
|
||||||
|
labs(title = 'No. of crashes in each country', x = '', y = '', fill = '') +
|
||||||
|
scale_fill_manual(values = map_pal, na.value = 'whitesmoke') +
|
||||||
|
theme(legend.position='right', legend.justification = "top") +
|
||||||
|
guides(fill = guide_legend(reverse = TRUE))
|
||||||
|
|
||||||
|
ggplotly(
|
||||||
|
p = ggplot_obj,
|
||||||
|
tooltip = "all",
|
||||||
|
dynamicTicks = FALSE,
|
||||||
|
layerData = 1,
|
||||||
|
originalData = TRUE
|
||||||
|
)
|
||||||
|
})
|
||||||
|
|
||||||
|
# poprawic szerkosc wierszy
|
||||||
|
output$allData <- DT::renderDataTable({
|
||||||
|
airplane_crashes %>%
|
||||||
|
# filter(Year == input$dataDates) %>%
|
||||||
|
DT::datatable()
|
||||||
|
})
|
||||||
|
|
||||||
|
}
|
||||||
|
|
||||||
|
shinyApp(ui = ui, server = server)
|
||||||
|
|
||||||
Loading…
Reference in New Issue
Block a user