TripPlanner.Rmd

---
title: "Trip Planner"
output: 
  flexdashboard::flex_dashboard:
    source_code: embed
runtime: shiny
---

```{r global, include=FALSE}
# Required Libraries
library(shiny)
library(leaflet)
library(sf)
library(TSP)
library(DT)

# Function to generate 20 random locations on Grand Bank of Newfoundland
generate_random_locations <- function(n = 20) {
  set.seed(123)
  # Approximate latitude and longitude boundaries of Grand Bank
  lat_range <- c(43.5, 46)
  lon_range <- c(-52, -49)
  
  # Generate random latitudes and longitudes within bounds
  data.frame(
    set = 1:n,
    latitude = runif(n, min = lat_range[1], max = lat_range[2]),
    longitude = runif(n, min = lon_range[1], max = lon_range[2])
  )
}

# Function to read CSV or generate random data
get_locations <- function(file) {
  if (is.null(file)) {
    generate_random_locations()
  } else {
    read.csv(file$datapath)
  }
}

# Function to convert lat/lon to UTM using sf
convert_to_utm <- function(data) {
  coordinates <- st_as_sf(data, coords = c("longitude", "latitude"), crs = 4326)
  utm_coords <- st_transform(coordinates, crs = 32621) # UTM Zone 21N for Newfoundland
  utm_coords_df <- as.data.frame(st_coordinates(utm_coords))
  return(cbind(data, utm_coords_df))
}

# Function to solve TSP
solve_tsp <- function(data, method = NULL) {
  dist_matrix <- dist(data[, c("X", "Y")])
  tsp <- TSP(dist_matrix)
  solution <- solve_TSP(tsp, method = method, control = list(start = 1))
  data[solution, ]
}
```

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

This application helps you plan efficient travel routes by suggesting the shortest distance path through a series of locations. The app applies [Traveling Salesman Problem (TSP)](https://en.wikipedia.org/wiki/Travelling_salesman_problem) algorithms to calculate and display the optimal route. See [Hahsler and Hornik (2007)](https://doi.org/10.18637/jss.v023.i02) for detailed methods.

---

Please upload a CSV file with the following columns: `set`, `latitude`, `longitude`.  
If no file is uploaded, random data will be generated.

```{r}
fileInput("file", "Upload CSV", accept = c(".csv"))
selectInput("method", "Select TSP method", 
            choices = c("nearest_insertion",
                        "farthest_insertion",
                        "cheapest_insertion",
                        "arbitrary_insertion",
                        "nn", 
                        "repetitive_nn",
                        "two_opt"),
            selected = "farthest_insertion")

get_locs <- reactive({
  file <- input$file
  locs <- get_locations(file)
  locs_utm <- convert_to_utm(locs)  # Convert to UTM
  tsp_solution <- solve_tsp(locs_utm, method = input$method)  # Solve TSP for optimal path
  # Calculate distances between consecutive points (nautical miles)
  tsp_solution$distance <- c(NA, round(sqrt(rowSums(diff(as.matrix(tsp_solution[, c("X", "Y")]))^2)), 2)) / 1000 * 0.539957
  tsp_solution
})

renderText({
  locs <- get_locs()
  paste("Total distance:", round(sum(locs$distance, na.rm = TRUE)), "nm")
})
```

Column
-----------------------------------------------------------------------

### Map of set locations and suggested path

```{r}
renderLeaflet({
  locs <- get_locs()
  leaflet(locs) %>%
    addTiles() %>%
    addCircleMarkers(
      lng = ~longitude, lat = ~latitude,
      radius = 2, color = "white", opacity = 1,
    ) %>%
    addLabelOnlyMarkers(
      lng = ~longitude, lat = ~latitude,
      label = ~as.character(set),
      labelOptions = labelOptions(noHide = TRUE, direction = "top", textOnly = TRUE, 
                                  style = list("color" = "white"))
    ) %>%
    addPolylines(
      lng = ~longitude, lat = ~latitude,
      color = "white", weight = 2, opacity = 0.8
    ) %>% 
    addProviderTiles(providers$Esri.WorldImagery, group="Two")
})
```

Column
-----------------------------------------------------------------------

### Set order and distances

```{r}
renderDT({
  locs <- get_locs()
  
  datatable(locs[, c("set", "latitude", "longitude", "distance")], 
            options = list(pageLength = 100, dom = "t", ordering = F),
            colnames = c("Set", "Latitude", "Longitude", "Distance (nm)"),
            rownames = FALSE, 
            editable = "cell") |> 
    formatRound(columns = c("latitude", "longitude", "distance"), digits = 3)
})
```