test.Rmd

---
title: "Corona Survey Tracker"
output: flexdashboard::flex_dashboard
runtime: shiny
---

Column {.sidebar}
-----------------------------------------------------------------------

```{r}
  selectInput("country", label = "Select country:",
          choices = c("Italy", "United States", "United Kingdom"), 
          selected = "Italy")
```

```{r setup, include=FALSE}
# Load library
library(flexdashboard)
library(limer)
library(kableExtra)
library(hrbrthemes)
library(dplyr)
library(tidyr)
library(shiny)
library(plotly)

# Setup API details
options(lime_api = 'https://survey3.gwdg.de/index.php?r=admin/remotecontrol')
options(lime_username = 'cimentada')
options(lime_password = 'BQ4J4VWpNdtQ')

# Do stuff with LimeSurvey API
get_session_key()  # Log in

autoInvalidate <- reactiveTimer(60000)

tst <- reactive({
  autoInvalidate()
  upd_time <<- paste("Data updated at", Sys.time())
  print(paste("Data updated at", upd_time))
  print(input$country)
  cnty_code <- switch(
    input$country,
    "Italy" = 684964,
    "United States" = 928449,
    "United Kingdom" = 558948
  )

  tst <- get_responses(cnty_code)
  tst$email_yes <- ifelse(tst$EMAIL == "", 0, 1)
  # Excludes email column
  tst <- tst[!names(tst) %in% "EMAIL"]
  tst$birth_groups <- cut(tst$BIRTH, breaks = c(0, 18, seq(20, 100, by = 10)))
  tst <- tst[, c("AREA", "birth_groups", "SEX", "email_yes")]
  
  tst <-
    tst %>%
    count(AREA, birth_groups, SEX, email_yes) %>%
    complete(AREA, birth_groups, SEX, email_yes, fill = list(n = 0)) %>%
    mutate(SEX = ifelse(SEX == "Prefer not to answer", "NA", SEX),
           birth_groups = gsub("\\(|\\]", "", birth_groups),
           birth_groups = gsub(",", "-", birth_groups),
           AREA = ifelse(AREA == "", "NA", AREA),
           age_sex = paste0(birth_groups, "-", SEX),
           perc = round(n / sum(n), 3),)
  tst
})

```

Column {data-width=650}
-----------------------------------------------------------------------

### Stratification tracker

```{r}
renderPlotly({
  plt <-
    tst() %>%
    group_by(AREA, age_sex) %>%
    summarize(n = sum(n)) %>% 
    ggplot(aes(x = age_sex, y = AREA, fill = n)) +
    geom_tile() +
    scale_x_discrete(guide = guide_axis(n.dodge = 2)) +
    scale_fill_viridis_c() +
    hrbrthemes::theme_ipsum() +
    theme(axis.text.x = element_text(angle = 90, hjust = 1))
  
  plt %>% 
    ggplotly()
})

```

Column {data-width=270}
-----------------------------------------------------------------------
### Update time

```{r}
renderText({
  autoInvalidate()
  upd_time
  })
```


### Total completed surveys

```{r}
renderPlotly({
  plt2 <-
    tst() %>%
    summarize(`Complete surveys` = sum(n),
              `Complete surveys w/ email` = sum(n[email_yes == 1])) %>%
    pivot_longer(cols = 1:2) %>% 
    ggplot(aes(name, value)) +
    geom_col() +
    hrbrthemes::theme_ipsum()

  plt2 %>% 
    ggplotly()
})



```