Hawkers.rmd

---
title: "Hawker, Voronois"
output: html_document
---

Clear 
```{r}
rm(list = ls())
```


```{r setup, include=FALSE}
library(devtools)
#devtools::install_github("JosiahParry/populartimes")

library(populartimes)
library(tidyverse)

#spatial
library(rgdal)
library(sp)
library(spdplyr)
library(raster)
library(terra)
library(sf)
library(rgeos)
```

Read Hawker Names + Location and Store in DF
```{r}
#read network
h = readOGR("Data/hawkers.mif", verbose = FALSE)
h = h %>% filter(STATUS != "Under Construction")

#filter and write back
writeOGR(obj=h, dsn="Data/hawkers_existing.mif", layer="00", driver="MapInfo File")

plot(h)

#save a copy
hh = h
```

Prepare New DF (only for popular times)
```{r}
h.df = data.frame(index = 1:nrow(hh),name = as.data.frame(hh)[,1] , lon = NA , lat = NA, popular_times = NA, time_spent = NA , current_popularity = NA, filled = NA )
h.df
```

Scrape: Receive Popular Times
```{r}
for (i in 1:nrow(h.df)){
#for (i in 1:1){
  hawker_name = h.df$name[i]
  #scrape data
  a = poptimes_from_address(hawker_name, "Singapore")
  
  #check poptimes if correct
  if (nrow(as.data.frame(a$popular_times)) != 0 ){ #a is correct
    #assign value
    h.df$popular_times[i] = a$popular_times
    h.df$lon[i] = a$lon
    h.df$lat[i] = a$lat
    h.df$time_spent[i] = a$time_spent
    h.df$current_popularity[i] = a$current_popularity
    
  } else {
    
    #check b (alternative query)
    b = poptimes_from_address(paste0("Hawker Centre ",hawker_name), paste0("Singapore"))
    if (nrow(as.data.frame(b$popular_times)) != 0 ){ #b is correct
      h.df$popular_times[i] = b$popular_times
      h.df$lon[i] = b$lon
      h.df$lat[i] = b$lat
      h.df$time_spent[i] = b$time_spent
      h.df$current_popularity[i] = b$current_popularity 
    }
    
    else {
      
      #check c (alternative query)
      c = poptimes_from_address(paste0("Hawker Centre"), paste0("Singapore ",hh$ADDRESSPOSTALCODE[i]))
      if (nrow(as.data.frame(c$popular_times)) != 0 ){ #c is correct
        h.df$popular_times[i] = c$popular_times
        h.df$lon[i] = c$lon
        h.df$lat[i] = c$lat
        h.df$time_spent[i] = c$time_spent
        h.df$current_popularity[i] = c$current_popularity
      
    }
     
  }
  
  }
  
}
missing = which(is.na(h.df$lat))
length(missing)
missing
h.df

as.data.frame(h.df)%>% filter(is.na(lon))
```

Manual Entry of Failed Queries
```{r}
h.df1 = h.df

inp = function(h.df, i, query1, query2){
  a = poptimes_from_address(query1, query2)
  h.df$popular_times[i] = a$popular_times
  h.df$lon[i] = a$lon
  h.df$lat[i] = a$lat
  h.df$time_spent[i] = a$time_spent
  h.df$current_popularity[i] = a$current_popularity
  return(h.df)
}

#if no point changing, just remove them completely,

h.df1 = inp(h.df1, 10, "Tanglin Halt Market", "48A Tanglin Halt Rd, Singapore 148813")  
h.df1 = inp(h.df1, 12, "Teban Gardens Food Centre", "37A Teban Gardens Rd, Singapore 601037") 
h.df1 = inp(h.df1, 15, "Telok Blangah Market", "82 Telok Blangah Dr, Singapore 100082")  #X
h.df1 = inp(h.df1, 17, "Hawker Centre", "30 Seng Poh Rd, #01-128/129, Singapore 168898")  #X
h.df1 = inp(h.df1, 23, "Lorong 8 Toa Payoh Hawker Centre", "210 Lor 8 Toa Payoh, Singapore 310210") 
h.df1 = inp(h.df1, 26, "Ayer Rajah Food Centre", "Blk 503, West Coast Drive. Ayer Rajah Food Centre, Singapore 120503") 
h.df1 = inp(h.df1, 28, "Hawker Centre", "91 Whampoa Dr, Singapore 320091") 
h.df1 = inp(h.df1, 33, "Bedok Reservoir Food Centre", "630 Bedok Reservoir Rd, Singapore 470630") 
h.df1 = inp(h.df1, 44, "Hawker Centre", "5 Cross St, Singapore 048418") #X
h.df1 = inp(h.df1, 47, "Maxwell Food Centre", "1 Kadayanallur St, Singapore 069184") 
h.df1 = inp(h.df1, 51, "Market", "Blk 58 New Upper Changi Rd, #01-07, Singapore 461058") 
h.df1 = inp(h.df1, 58, "Albert Centre", "270 Queen St, Singapore 180270") 
h.df1 = inp(h.df1, 69, "Ang Mo Kio Central Market & Food Centre", "724 Ang Mo Kio Ave 6, Singapore 560724") 
h.df1 = inp(h.df1, 71, "Bedok Food Centre", "1 Bedok Rd, Singapore 469572") 
h.df1 = inp(h.df1, 81, "Circuit Road Market Food Centre", "80 Circuit Rd, Singapore 370080") 
h.df1 = inp(h.df1, 83, "MacPherson Market & Food Centre", "89 Circuit Rd, Singapore 370089") 
h.df1 = inp(h.df1, 84, "Hawker Centre", "353 Clementi Ave 2, Singapore 120353")  #X
h.df1 = inp(h.df1, 85, "Clementi 448 Market & Food Centre", "448 Clementi Ave 3, Singapore 120448") 
h.df1 = inp(h.df1, 86, "West Coast Food Centre", "726 Clementi West Street 2, Singapore 120726") 
h.df1 = inp(h.df1, 91, "Market", "7 Empress Rd, Singapore 260007") 
h.df1 = inp(h.df1, 95, "Yuan Authentic Thai Stewed Beef Noodles", "20 Ghim Moh Road #01-32, Ghim Moh Market, & Food Center, 270020") 
h.df1 = inp(h.df1, 102, "Hussain Muslim Food", "209 Hougang Street 21, #01-54, Singapore 530209") 



h.df1

#test query
as.data.frame(h.df1[20,]$popular_times)

#check for empty entries
as.data.frame(h.df1)%>% filter(is.na(lon))
```

Manual Entry of Failed Queries: Check if any hawker centre has no popular times (15,17,44,58,84)
```{r}
a = c()
for (i in 1:nrow(h.df1)){
  #check if filled
  if ( nrow(as.data.frame(h.df1$popular_times[i])) == 0 ){
    a= c(a,i)
  }
}
print(a)
```

Get the peak times (Mode)
```{r}
#formula for mode
Mode <- function(x) {
  ux <- unique(x)
  ux[which.max(tabulate(match(x, ux)))]
}

h.df1$peak = NA

for (i in 1:nrow(h.df1)){
  #if have data
  if ( nrow(as.data.frame(h.df1$popular_times[i])) > 1 ){
    t = as.data.frame(h.df1[i,]$popular_times) %>% arrange(desc(popularity)) %>% .[1:10,] %>%.$hour 
    h.df1$peak[i] = Mode(t)
  }
}
h.df1 %>% arrange(peak)
```

Upload Peak Data on Voronoi Map
```{r}
h.df1
v = readOGR("Data/hawkers_v_clean.mif", verbose = FALSE)
v$peak = h.df1$peak

#filter and write back
writeOGR(obj=v, dsn="Data/hawkers_v_peak.mif", layer="00", driver="MapInfo File")
```