weather.Rmd

---
title: "MPG Ranch Weather"
output:
  flexdashboard::flex_dashboard: 
    theme: spacelab
runtime: shiny
---

```{r setup, include=FALSE}
library(flexdashboard)
library(shiny)
library(plotly)
library(lubridate)
library(readr)
library(DT)
library(leaflet)
library(colorspace)
```

```{r load_data, include = FALSE}
weather_df <- read_csv('https://storage.googleapis.com/mpg-data-warehouse/weather_summaries/mpg_ranch_daily/weather_daily_cache.csv')

stations_df <- read_csv('https://storage.googleapis.com/mpg-data-warehouse/weather/MPG/station_meta/weather_station_location.csv')

pal <- sequential_hcl(palette = "Batlow", n = 6)

station_cols <- scale_color_manual(
  name = "Stations:",
  values = c("Baldy Draw" = pal[1], 
             "Baldy Summit" = pal[2], 
             "Indian Ridge" = pal[3], 
             "Orchard House" = pal[4], 
             "Sanfoin Bench" = pal[5], 
             "South Baldy Ridge" = pal[6])
)

station_fills <- scale_fill_manual(
  name = "Stations",
  values = c("Baldy Draw" = pal[1], 
             "Baldy Summit" = pal[2], 
             "Indian Ridge" = pal[3], 
             "Orchard House" = pal[4], 
             "Sanfoin Bench" = pal[5], 
             "South Baldy Ridge" = pal[6])
)

theme_text_col = "gray35"
theme_weather =
  theme_bw() +
  theme(
    axis.title.y = element_text(size = 12, color = theme_text_col, margin = unit(c(t = 0, r = 4, b = 0, l = 0 ), "mm")),
    axis.text.x = element_text(size = 10, color = theme_text_col, margin = unit(c(t = 4, r = 0, b = 0, l = 0 ), "mm")),
    axis.text.y = element_text(size = 10, color = theme_text_col, margin = unit(c(t = 0, r = 3, b = 0, l = 0 ), "mm")),
    axis.ticks.length = unit(-1.5, "mm"),
    strip.text = element_text(size = 12, color = "gray15"),
    strip.background = element_rect(fill = "lightsteelblue2"),
    plot.caption = element_text(size = 10, color = theme_text_col),
    legend.text = element_text(size = 12, color = theme_text_col), 
    legend.title = element_text(size = 12, color = theme_text_col),
    legend.position = "top", 
    panel.grid.minor = element_blank()
  )
```

Sidebar {.sidebar data-width=250}
=====================================
```{r}

dateRangeInput("date_range", 
               label = "Date range:",
               start = "2013-01-01",
               end = Sys.Date() - 1,
               max = Sys.Date() - 1,
               min = "2013-01-01")

selectInput("station",
            "Select station:",
            choices = c("Baldy Draw", "Baldy Summit", "Indian Ridge",
                        "Orchard House", "Sanfoin Bench", "South Baldy Ridge"),
            selected = c("Baldy Summit", "Orchard House"),
            multiple = TRUE)

sliderInput("day_slider", 
            label = "Day Range by Year:",
            min = 1, 
            max = 366, 
            value = c(1,366), 
            step = 1)

show_weather_df <- reactive({
  weather_df %>%
    select(-X1) %>%
    filter(date_day >= min(input$date_range) &
           date_day <= max(input$date_range)) %>%
    filter(station %in% if(is.null(input$station)) {'Orchard House'} else {input$station})  %>%
    mutate(doy = yday(date_day)) %>%
    filter(doy >= min(input$day_slider) &
           doy <= max(input$day_slider))
        
})

show_stations_df <- reactive({
  stations_df %>%
    filter(station %in% input$station)
})

# Create placeholder for the download Button
uiOutput("downloadUI")

# Create the actual download Button
output$downloadUI <- renderUI({
  downloadButton("downBtn", "Download *.csv", style = "width:100%;")
})

# Add download handling
output$downBtn <- downloadHandler(
  filename = function() {
    paste('mpg_weather-', min(input$date_range), '_', max(input$date_range), '.csv', sep='')
  },
  content = function(file) {
    write.csv(show_weather_df(), file, row.names = FALSE)
  }
)
```

Plots
=====================================

Column {.tabset data-width=750}
-----------------
### Temperature

```{r}

renderPlot({
  show_weather_df() %>%
    select(date_day, station, temp_F_mean) %>%
    arrange(date_day) %>%
    mutate(
      year = year(date_day), 
      doy = yday(date_day),
      axis_date = as.Date(paste0("2000-",format(date_day, "%j")), "%Y-%j")
      ) %>%
    ggplot(aes(x=axis_date, y=temp_F_mean)) +
    geom_line(aes(colour = station), size = 0.8) +
    facet_grid(vars(year), as.table = FALSE) +
    labs(x = "", y = expression(Average~Temperature~"("*degree*F*")")) +
    scale_x_date(labels = function(x) format(x, "%b-%d")) +
    station_cols +
    theme_weather 
})

```

### Precipitation

```{r}
renderPlot({
  show_weather_df() %>%
    arrange(date_day) %>%
    mutate(
      year = year(date_day), 
      doy = yday(date_day),
      axis_date = as.Date(paste0("2000-",format(date_day, "%j")), "%Y-%j")
      ) %>%
    group_by(year, station) %>%
    mutate(precip_in = cumsum(precip_in_sum)) %>%
    ungroup() %>%
    ggplot(aes(x = axis_date, y = precip_in, group = station)) +
    geom_step(aes(color = station), size = 0.8) +
    facet_grid(vars(year), as.table = FALSE) +
    labs(x = "", y = "Cumulative Rainfall (in)", caption = "Note: the weather stations do not measure snow-water-equivalent. Values shown here depict rainfall, with trivial contributions from snow.") +
    scale_x_date(labels = function(x) format(x, "%b-%d")) +
    station_cols +
    theme_weather
})
```

### Wind Speed

```{r}
renderPlot({
  show_weather_df() %>%
    select(date_day, station, wspd_mph_mean) %>%
    arrange(date_day) %>%
    mutate(
      year = year(date_day), 
      doy = yday(date_day),
      axis_date = as.Date(paste0("2000-",format(date_day, "%j")), "%Y-%j")
      ) %>%
    ggplot(aes(x=axis_date, y=wspd_mph_mean)) +
    geom_line(aes(colour = station), size = 0.8) +
    facet_grid(vars(year), as.table = FALSE) +
    labs(x = "", y = "Average Wind Speed (mph)") +
    scale_x_date(labels = function(x) format(x, "%b-%d")) +
    station_cols +
    theme_weather
})
```

### Wind Direction

```{r}
ordinal_directions <- c('N', 'E', 'S', 'W', 'N')

renderPlot({
  show_weather_df() %>%
    select(date_day, station, wdir_deg_prevail) %>%
    arrange(date_day) %>%
    mutate(
      year = year(date_day), 
      doy = yday(date_day),
      axis_date = as.Date(paste0("2000-",format(date_day, "%j")), "%Y-%j")
      ) %>%
    ggplot(aes(x=axis_date, y=wdir_deg_prevail, group = station)) +
    geom_point(aes(fill = station), shape = 21, color = "black", stroke = 0.4, alpha = 0.7, size = 1.6) +
    facet_grid(vars(year), as.table = FALSE) +
    labs(x = "", y = "Prevailing Wind Direction") +
    scale_x_date(labels = function(x) format(x, "%b-%d")) +
    scale_y_continuous(labels = ordinal_directions, breaks = (0:4) * 90, limits = c(0, 360)) +
    station_fills +
    theme_weather
})
```


Table
=====================================

```{r}
renderDataTable({
  datatable(
    show_weather_df() %>% select(-doy), 
    rownames = FALSE,
    options = list(scrollY = '400px', scrollX = TRUE)
  )
})
```

Map
=====================================

```{r}
leaflet() %>%
#  addTiles() %>%
  setView(-114.001, 46.70, zoom = 13) %>%
  addWMSTiles("https://server.arcgisonline.com/ArcGIS/rest/services/World_Topo_Map/MapServer/tile/{z}/{y}/{x}",
              layers = "OpenTopoMap") %>%
  addMarkers(lat = stations_df$lat, lng = stations_df$long, 
             popup = paste0(popup = paste0(
                                      "<strong>Station: </strong>", stations_df$station, "<br>",
                                      "<strong>Elevation: </strong>", stations_df$elevation, " ft")),
             label = stations_df$station)
```

About
=====================================
#### Intention
MPG Ranch has tools which display weather data, but none of them allow general users to download data for their own use. In this application, users may filter, view, and *download* weather data from MPG Ranch weather stations.

#### Description
The data are collected from six weather stations, ranging from 3218 to 5994 ft in elevation. The stations record weather variables every 30 minutes. These observations are then loaded into the MPG Ranch Data Warehouse and compiled into daily values. Data included here begin on January 1, 2013, and new data are appended daily.

Leap years have 366 days. Users should account for this when processing downloaded data. 

Known missing data are displayed [here](https://datastudio.google.com/reporting/c55060be-c84b-4547-8556-5107f9a51813){target="_blank"}.

Thank you Nick and Gus for setting up and maintaining the weather stations.

#### Feedback
We appreciate suggestions. Please contact Beau and Erik with any feedback. If you would like to report bugs or issues, or would like to be involved with development, join us on [GitHub](https://github.com/samsoe/mpg_shiny/issues){target="_blank"}.