analysis-stationAreaAccessPlan.Rmd

---
title: "Analysis: Station Area Access Plans"
site: distill::distill_website
output:
  distill::distill_article:
    toc: false
    toc_depth: 4
    toc_float: true
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = FALSE)

# Learn more about creating websites with Distill at:
# https://rstudio.github.io/distill/website.html

# Learn more about publishing to GitHub Pages at:
# https://rstudio.github.io/distill/publish_website.html#github-pages

library(mapboxapi)
library(osmdata)
library(sf)
library(sfnetworks)
library(tidyverse)
library(tmap)

tmap_options(check.and.fix = TRUE)
tmap_mode("view")

```

```{css, echo=FALSE}

d-title {
  display:none;
}

```

<!-- Use a HTML break to add a little white space between the title and the map -->
<br>

<!-- Add a disclaimer that this is still being edited. -->

<!-- <div class="boxed"> -->
<!--   This website is still a work-in-progress and may not show accurate or complete information! -->
<!-- </div>  -->

# Metro Station Area Access Plans (SAAP)

<p style="color:#e85d04;">**Status**: Analysis in Progress as of `r format(Sys.Date(), "%d %B, %Y")`</p>

## What the CMP Says

"*Metro Station Area Access Plans (SAAP) are proposed to reduce the requirement for commuters to bring their personal vehicle or use public transport for their commute to and from metro stations. For each station, area of 1km radius will be covered under the SAAP to improve the road network to TenderSure specifications. It shall also assess the demand for park-and-ride, the last mile destinations (and distances), currently available last mile mode options etc. SAAP shall propose plans to integrate required last mile modes within the station area for passenger convenience. The SAAP shall also propose critical intervention areas to improve safety of pedestrians in station areas, identify bottleneck locations and suggest solutions to ease traffic movement. To begin with 4 terminal stations and 11 other stations in areas having high population density are proposed for implementing SAAP, and thereafter other stations are to be covered.*" (BMRCL et al., 2020, p. 4-42)

CMP 2020 sees the SAAP projects as a way to meet the **NUTP Objective** to "*Encourage greater use of public transport and non-motorized modes*". (BMRCL et al., 2020, p. 4-42)

## The Stations

The stations identified are given in Table 4-14 reproduced below. (BMRCL et al., 2020, p-4-42)

Sl. No.| Metro Station
-|-
1. | Baiyapannahalli Station
2. | Yellachanahalli
3. | Nayandanahalli
4. | Nagasandra
5. | Sampige Mantri Square
6. | Sriramapura
7. | Rajajinagar
8. | Mahakavi Kuvempu Road
9. | Mahalakshmi
10. | Sandal Soap Factory
11. | Yeshwantpur
12. | Peenya
13. | Peenya Industry
14. | Dasarahalli

The map below identifies only these stations. Other stations are not shown.

```{r}
# First, I will set up a base map with the BBMP ward boundaries.

bangaloreWardBoundary <- read_sf(here::here("data/raw-data/bangaloreWardBoundary.shp"))

# Transform it for consistency

bangaloreWardBoundary <- bangaloreWardBoundary %>% 
  st_transform(3857)

# Now load in the dataset for this map. To start with, its going to have just points.

bangaloreCMPProjects <- readxl::read_xlsx(here::here("data/raw-data/bangaloreCMP2020Projects.xlsx"))

# Filter out the locations that are not geo-located yet

bangaloreCMPProjects <- bangaloreCMPProjects %>% 
  filter(!is.na(location)) %>% 
  separate(location, sep = ", ", into = c("lon", "lat"))

# Convert lat and long into geometry

bangaloreCMPProjects <- bangaloreCMPProjects %>% 
  st_as_sf(coords = c("lat", "lon"), crs = 4326) %>% 
  st_transform(3857)

# Define a Palette

myPalette <- c("#f94144",
               "#f3722c",
               "#f8961e",
               "#f9844a",
               "#f9c74f",
               "#90be6d",
               "#43aa8b",
               "#4d908e",
               "#577590",
               "#277da1")

```

```{r}
# BASE MAP

# I want the base map to show the metro lines, the stations included in the SAAP and the ones that are excluded.

# I'll get this data from Open Street Maps

# # Purple Line
# 
# query <- getbb("Bangalore") %>% 
#   opq() %>%
#   add_osm_feature("wikidata", "Q7261433")
# 
# str(query)
# 
# purpleLine <- osmdata_sf(query)
# 
# purpleLine <- purpleLine$osm_lines
# 
# purpleLine <- purpleLine %>% 
#   select(osm_id, geometry) %>% 
#   mutate(color = "#e542de")
# 
# # Green Line
# 
# query <- getbb("Bangalore") %>% 
#   opq() %>%
#   add_osm_feature("wikidata", "Q17054000")
# 
# str(query)
# 
# greenLine <- osmdata_sf(query)
# 
# greenLine <- greenLine$osm_lines
# 
# greenLine <- greenLine %>% 
#   select(osm_id, geometry) %>% 
#   mutate(color = "#009933")
# 
# metroRoutes <- bind_rows(purpleLine, greenLine)
# 
# saveRDS(metroRoutes, here::here("data/raw-data/metroRoutes.rds"))

metroRoutes <- readRDS(here::here("data/raw-data/metroroutes.rds"))

metroRoutes <- metroRoutes %>% 
  st_transform(3857)

# I'll draw the base map here.

mapBase <- tm_shape(bangaloreWardBoundary) +
  tm_borders(col = "#000000",
             lwd = 2)
  
mapMetroRoutes <- tm_shape(metroRoutes) +
  tm_lines(col = "color",
           popup.vars = FALSE)

```

```{r}
# I also want the metro stations to be more prominent on the maps

# Station Buildings
# 
# query <- getbb("Bangalore") %>%
#   opq() %>%
#   add_osm_feature("building", "train_station")
# 
# str(query)
# 
# bangaloreStationBuildings <- osmdata_sf(query)
# 
# bangaloreStationBuildings <- bangaloreStationBuildings$osm_polygons
# 
# saveRDS(bangaloreStationBuildings, here::here("data/raw-data/bangaloreStationBuildings.rds"))

bangaloreStationBuildings <- readRDS(here::here("data/raw-data/bangaloreStationBuildings.rds"))

bangaloreStationBuildings <- bangaloreStationBuildings %>% 
  select(osm_id, geometry) %>% 
  st_transform(3857)

```

```{r}
# STATION AREA ACCESS PLAN PROJECTS

# Filter only the station area access plan projects

stationAreaAccess <- bangaloreCMPProjects %>% 
  filter(type == "Station Area Access Plan")

# Rename columns for neater display

stationAreaAccess <- stationAreaAccess %>%
  rename(ID = "id",
         Type = "type",
         Description = "description_0")

# Rearrange columns so that I get the description in the pop-up

stationAreaAccess <- stationAreaAccess %>% 
  select(Description, ID, Type, geometry)

# Add correct colours to the stations

stationAreaAccess <- stationAreaAccess %>% 
  mutate(color = case_when(Description == "Baiyapannahalli Station" ~ "#e542de",
                           Description == "Nayandanahalli Station" ~ "#e542de",
                           TRUE ~ "#009933"))

# Draw Junction Improvement Map with Buffer

mapStationAreaAccess <- stationAreaAccess %>% 
  tm_shape() +
  tm_dots(col = "color",
          popup.vars = c("ID", "Type", "Description"))

# stationAreaAccess %>%
#   st_buffer(300) %>% 
#   st_intersection(bangaloreStationBuildings) %>% 
#   tm_shape() +
#   tm_fill(col = "color",
#           popup.vars = c("ID", "Type", "Description"))

```

```{r, layout="l-screen-inset", fig.height=6.75}
# Publish combined map

mapBase + mapMetroRoutes + mapStationAreaAccess
```

## Analysis

The map below shows the area that will be considered as part of the SAAP projects, that is, a one-kilometer radius around each station.

```{r, layout="l-screen-inset", fig.height=6.75}
# Draw a 1km buffer around the stations

stationAreaAccessBuffers <- stationAreaAccess %>% 
  st_buffer(1000)

stationAreaAccessBuffersUnion <- st_union(stationAreaAccessBuffers)

# Include these buffers on the map

mapStationAreaBuffers <- stationAreaAccessBuffers %>% 
  tm_shape() +
  tm_borders(col = "#000000",
             lty = "dashed",
             lwd = 1.2)

mapBase + mapMetroRoutes + mapStationAreaAccess + mapStationAreaBuffers
```

The description of the SAAP projects defines the first objective as reducing the need for commuters to bring their vehicles or use public transport in their commute to and from metro stations. This leaves on-demand shared mobility services, shuttles, and walking as options. For now, let's focus on walking. It is generally accepted that people are comfortable walking 500m. For an adult, this would take about 6 minutes. While a circle with a radius of 500m is often drawn to indicate the walkable distance, it is more useful to look at a route-based walkable distance; this helps exclude private property and inaccessible areas. The map below identifies these walkable areas. Additionally, areas accessible in 10 minutes and 15 minutes (the-fifteen-minute-city) are also shown.

```{r, layout="l-screen-inset", fig.height=6.75}

# I will calculate walking isochrones for each metro station.

# 
# walkableIsochrones <- mb_isochrone(stationAreaAccess,
#                                    profile = "walking",
#                                    time = c(6, 10, 15),
#                                    access_token = myToken)
# 
# saveRDS(walkableIsochrones, here::here("data/raw-data/walkableIsochrones.rds"))

walkableIsochrones <- readRDS(here::here("data/raw-data/walkableIsochrones.rds"))

# Transform for consistency

walkableIsochrones <- walkableIsochrones %>% 
  st_transform(3857)

mapWalkingTime <- tm_shape(walkableIsochrones) +
  tm_fill(col = "time",
          palette = "-YlOrBr",
          alpha = 0.5,
          breaks = c(6, 10, 15, 20),
          labels = c("Upto 6 minutes",
                     "Between 6 and 10 minutes",
                     "Between 10 and 15 minutes"),
          title = "Walking Time",
          popup.vars = FALSE)

mapBase + mapWalkingTime + mapMetroRoutes + mapStationAreaAccess + mapStationAreaBuffers
```

Although patterns begin to emerge from this map, it would be more useful to focus on just the roads. Further, it is interesting to note that the one-kilometer radius buffers and walkable areas overlap for many of the metro stations. Development within these overlapped areas would benefit multiple stations and would be an efficient use of resources. Keeping this in mind, it would also make sense to align the analysis accordingly and divide the metro stations into groups, as indicated in the table below. 

Use the links in the table to navigate to specific sections.

Link | Stations Included
-|-
[Group 1](#group-1) | Nagasandra Station, Dasarahalli Station, Peenya Industry Station, Peenya Station
[Group 2](#group-2) | Yeshwantpur Station, Sandal Soap Factory Station,  Mahalakshmi Station, Rajajinagar Station, Mahakavi Kuvempu Road Station,  Sriramapura Station, Sampige Mantri Square Station
[Group 3](#group-3) | Nayandanahalli Station
[Group 4](#group-4) | Yellachanahalli Station
[Group 5](#group-5) | Baiyapannahalli Station

```{r, layout="l-screen-inset", fig.height=6.75}

# Load in the roads

bangaloreRoads <- readRDS(here::here("data/raw-data/bangaloreRoads.rds"))

bangaloreRoads <- bangaloreRoads %>% 
  st_transform(3857)

# Add the grouping to the stations

stationAreaAccess <- stationAreaAccess %>% 
  mutate(Group = case_when(Description == "Baiyapannahalli Station" ~ 5,
                            Description == "Yellachanahalli Station" ~ 4,
                            Description == "Nayandanahalli Station" ~ 3,
                            Description == "Nagasandra Station" ~ 1,
                            Description == "Dasarahalli Station" ~ 1,
                            Description == "Peenya Industry Station" ~ 1,
                            Description == "Peenya Station" ~ 1,
                            TRUE ~ 2))

# Add the grouping to the stations buffers

stationAreaAccessBuffers <- stationAreaAccessBuffers %>% 
  mutate(Group = case_when(Description == "Baiyapannahalli Station" ~ 5,
                            Description == "Yellachanahalli Station" ~ 4,
                            Description == "Nayandanahalli Station" ~ 3,
                            Description == "Nagasandra Station" ~ 1,
                            Description == "Dasarahalli Station" ~ 1,
                            Description == "Peenya Industry Station" ~ 1,
                            Description == "Peenya Station" ~ 1,
                            TRUE ~ 2))

# Intersect the roads with the buffer areas

bangaloreRoadsBuffer <- st_intersection(bangaloreRoads, stationAreaAccessBuffers)

# Intersect these roads with the walking time isochrones

bangaloreRoadsIsochrone <- st_intersection(bangaloreRoadsBuffer, walkableIsochrones)

```

```{r}
# Load data and wrangle it. I'll look at how far it is to walk to bus stations from each of the metro stations.

bangaloreBusStops <- readRDS(here::here("data/raw-data/bangaloreBusStops.rds"))

bangaloreBusStops <- bangaloreBusStops$osm_points

bangaloreBusStops <- bangaloreBusStops %>% 
  select(osm_id, name, geometry) %>% 
  rename(OSM_ID = "osm_id",
         Name = "name") %>% 
  st_transform(3857)
```

### Group 1 {#group-1}

```{r, layout="l-screen-inset", fig.height=6.75}
# Draw the maps out

mapStationAreaAccessG1 <- stationAreaAccess %>% 
  filter(Group == 1) %>% 
  tm_shape() +
  tm_symbols(col = "color",
             border.col = "#000000",
             border.lwd = 4,
             popup.vars = c("ID", "Type", "Description"))

mapStationAreaBuffersG1 <- stationAreaAccessBuffers %>% 
  filter(Group == 1) %>% 
  tm_shape() +
  tm_borders(col = "#000000",
             lty = "dashed",
             lwd = 1.2)

mapWalkingTimeG1 <- bangaloreRoadsIsochrone %>% 
  filter(Group == 1) %>% 
  tm_shape() +
  tm_lines(col = "time",
           lwd = 2,
          palette = "-YlOrBr",
          breaks = c(6, 10, 15, 20),
          labels = c("Upto 6 minutes",
                     "Between 6 and 10 minutes",
                     "Between 10 and 15 minutes"),
          title.col = "Walking Time",
          popup.vars = FALSE)

mapStationAreaBuffersG1 + mapWalkingTimeG1 + mapStationAreaAccessG1 + mapMetroRoutes

```

Within the one-kilometer radius buffers are a number of bus stops. The map below shows the location of these stops and the time it would take to walk to the metro station from them. Three of the four stations have bus stops within a 6 minute walk of them.

```{r}
# Overlay bus stops on these maps

mapBusStopsG1 <- stationAreaAccessBuffers %>% 
  filter(Group == 1) %>%
  st_intersection(bangaloreBusStops) %>% 
  select(Name, OSM_ID, geometry) %>% 
  tm_shape() +
  tm_symbols(shape = 21, 
             col = "#880d1e",
             border.col = "#000000",
             border.lwd = 3,
             size = 0.2,
             popup.vars = c("OSM_ID", "Name"))

```

```{r, layout="l-screen-inset", fig.height=6.75}
# Map with bus stops

mapStationAreaBuffersG1 + mapWalkingTimeG1 + mapStationAreaAccessG1 + mapMetroRoutes + mapBusStopsG1

```

### Group 2 {#group-2}

```{r, layout="l-screen-inset", fig.height=6.75}
# Draw the maps out

mapStationAreaAccessG2 <- stationAreaAccess %>% 
  filter(Group == 2) %>% 
  tm_shape() +
  tm_symbols(col = "color",
             border.col = "#000000",
             border.lwd = 4,
             popup.vars = c("ID", "Type", "Description"))

mapStationAreaBuffersG2 <- stationAreaAccessBuffers %>% 
  filter(Group == 2) %>% 
  tm_shape() +
  tm_borders(col = "#000000",
             lty = "dashed",
             lwd = 1.2)

mapWalkingTimeG2 <- bangaloreRoadsIsochrone %>% 
  filter(Group == 2) %>% 
  tm_shape() +
  tm_lines(col = "time",
           lwd = 2,
          palette = "-YlOrBr",
          breaks = c(6, 10, 15, 20),
          labels = c("Upto 6 minutes",
                     "Between 6 and 10 minutes",
                     "Between 10 and 15 minutes"),
          title.col = "Walking Time",
          popup.vars = FALSE)

mapStationAreaBuffersG2 + mapWalkingTimeG2 + mapStationAreaAccessG2 + mapMetroRoutes

```

```{r}
# Overlay bus stops on these maps

mapBusStopsG2 <- stationAreaAccessBuffers %>% 
  filter(Group == 2) %>%
  st_intersection(bangaloreBusStops) %>% 
  select(Name, OSM_ID, geometry) %>% 
  tm_shape() +
  tm_symbols(col = "#880d1e",
             border.col = "#000000",
             border.lwd = 3,
             size = 0.2,
             popup.vars = c("OSM_ID", "Name"))

```

```{r, layout="l-screen-inset", fig.height=6.75}
# Map with bus stops

mapStationAreaBuffersG2 + mapWalkingTimeG2 + mapStationAreaAccessG2 + mapMetroRoutes + mapBusStopsG2

```

### Group 3 {#group-3}

```{r, layout="l-screen-inset", fig.height=6.75}
# Draw the maps out

mapStationAreaAccessG3 <- stationAreaAccess %>% 
  filter(Group == 3) %>% 
  tm_shape() +
  tm_symbols(col = "color",
             border.col = "#000000",
             border.lwd = 4,
             popup.vars = c("ID", "Type", "Description"))

mapStationAreaBuffersG3 <- stationAreaAccessBuffers %>% 
  filter(Group == 3) %>% 
  tm_shape() +
  tm_borders(col = "#000000",
             lty = "dashed",
             lwd = 1.2)

mapWalkingTimeG3 <- bangaloreRoadsIsochrone %>% 
  filter(Group == 3) %>% 
  tm_shape() +
  tm_lines(col = "time",
           lwd = 2,
          palette = "-YlOrBr",
          breaks = c(6, 10, 15, 20),
          labels = c("Upto 6 minutes",
                     "Between 6 and 10 minutes",
                     "Between 10 and 15 minutes"),
          title.col = "Walking Time",
          popup.vars = FALSE)

mapWalkingTimeG3 + mapStationAreaBuffersG3 + mapStationAreaAccessG3 + mapMetroRoutes

```

```{r}
# Overlay bus stops on these maps

mapBusStopsG3 <- stationAreaAccessBuffers %>% 
  filter(Group == 3) %>%
  st_intersection(bangaloreBusStops) %>% 
  select(Name, OSM_ID, geometry) %>% 
  tm_shape() +
  tm_symbols(col = "#880d1e",
             border.col = "#000000",
             border.lwd = 3,
             size = 0.2,
             popup.vars = c("OSM_ID", "Name"))

```

```{r, layout="l-screen-inset", fig.height=6.75}
# Map with bus stops

mapStationAreaBuffersG3 + mapWalkingTimeG3 + mapStationAreaAccessG3 + mapMetroRoutes + mapBusStopsG3

```

### Group 4 {#group-4}

```{r, layout="l-screen-inset", fig.height=6.75}
# Draw the maps out

mapStationAreaAccessG4 <- stationAreaAccess %>% 
  filter(Group == 4) %>% 
  tm_shape() +
  tm_symbols(col = "color",
             border.col = "#000000",
             border.lwd = 4,
             popup.vars = c("ID", "Type", "Description"))

mapStationAreaBuffersG4 <- stationAreaAccessBuffers %>% 
  filter(Group == 4) %>% 
  tm_shape() +
  tm_borders(col = "#000000",
             lty = "dashed",
             lwd = 1.2)

mapWalkingTimeG4 <- bangaloreRoadsIsochrone %>% 
  filter(Group == 4) %>% 
  tm_shape() +
  tm_lines(col = "time",
           lwd = 2,
          palette = "-YlOrBr",
          breaks = c(6, 10, 15, 20),
          labels = c("Upto 6 minutes",
                     "Between 6 and 10 minutes",
                     "Between 10 and 15 minutes"),
          title.col = "Walking Time",
          popup.vars = FALSE)

mapStationAreaBuffersG4 + mapWalkingTimeG4 + mapStationAreaAccessG4 + mapMetroRoutes

```

```{r}
# Overlay bus stops on these maps

mapBusStopsG4 <- stationAreaAccessBuffers %>% 
  filter(Group == 4) %>%
  st_intersection(bangaloreBusStops) %>% 
  select(Name, OSM_ID, geometry) %>% 
  tm_shape() +
  tm_symbols(col = "#880d1e",
             border.col = "#000000",
             border.lwd = 3,
             size = 0.2,
             popup.vars = c("OSM_ID", "Name"))

```

```{r, layout="l-screen-inset", fig.height=6.75}
# Map with bus stops

mapStationAreaBuffersG4 + mapWalkingTimeG4 + mapStationAreaAccessG4 + mapMetroRoutes + mapBusStopsG4

```

### Group 5 {#group-5}

```{r, layout="l-screen-inset", fig.height=6.75}
# Draw the maps out

mapStationAreaAccessG5 <- stationAreaAccess %>% 
  filter(Group == 5) %>% 
  tm_shape() +
  tm_symbols(col = "color",
             border.col = "#000000",
             border.lwd = 4,
             popup.vars = c("ID", "Type", "Description"))

mapStationAreaBuffersG5 <- stationAreaAccessBuffers %>% 
  filter(Group == 5) %>% 
  tm_shape() +
  tm_borders(col = "#000000",
             lty = "dashed",
             lwd = 1.2)

mapWalkingTimeG5 <- bangaloreRoadsIsochrone %>% 
  filter(Group == 5) %>% 
  tm_shape() +
  tm_lines(col = "time",
           lwd = 2,
          palette = "-YlOrBr",
          breaks = c(6, 10, 15, 20),
          labels = c("Upto 6 minutes",
                     "Between 6 and 10 minutes",
                     "Between 10 and 15 minutes"),
          title.col = "Walking Time",
          popup.vars = FALSE)

mapStationAreaBuffersG5 + mapWalkingTimeG5 + mapStationAreaAccessG5 + mapMetroRoutes

```

```{r}
# Overlay bus stops on these maps

mapBusStopsG5 <- stationAreaAccessBuffers %>% 
  filter(Group == 5) %>%
  st_intersection(bangaloreBusStops) %>%
  select(Name, OSM_ID, geometry) %>% 
  tm_shape() +
  tm_symbols(col = "#880d1e",
             border.col = "#000000",
             border.lwd = 3,
             size = 0.2,
             popup.vars = c("OSM_ID", "Name"))

```

```{r, layout="l-screen-inset", fig.height=6.75}
# Map with bus stops

mapStationAreaBuffersG5 + mapWalkingTimeG5 + mapStationAreaAccessG5 + mapMetroRoutes + mapBusStopsG5

```

## Impact of other CMP 2020 Proposals (TBD)

Other projects from the CMP 2020 that are within these one-kilometer radius buffer areas are shown in the map below. (Note: This will change as the geo-referencing process continues. More projects *may* be included.)

```{r, layout="l-screen-inset", fig.height=6.75}
# Here, I want to display the other projects within this buffer area.

otherProposals <- bangaloreCMPProjects %>% 
  filter(type != "Station Area Access Plan")

otherProposals <- st_intersection(otherProposals, stationAreaAccessBuffersUnion)

# Then I want some amount of info to be displayed when I click on it. I will restrict the info shown to the ID, name and project type.

otherProposals <- otherProposals %>% 
  rename(Description = "description_0",
         ID = "id",
         Type = "type")

otherProposals <- otherProposals %>% 
  select(Description, ID, Type, geometry)

mapOtherProposals <- otherProposals %>% 
  tm_shape() +
  tm_dots(col = "#000000",
          popup.vars = c("ID", "Type", "Description"))

mapBase + mapMetroRoutes + mapStationAreaAccess + mapStationAreaBuffers + mapOtherProposals
```