IntroToMapping.Rmd

---
title: "Getting Started with Geographic Data Displays using R"
author: | 
  | Vivek H. Patil
  | Associate Professor of Marketing, Gonzaga University
date: "August 14, 2015"
output:
  ioslides_presentation:
    smaller: yes
    theme: spacelab
    widescreen: yes
---

<style>
.title-slide hgroup h1 {color: red;}
h2 {color: red;}
</style>

```{r setup, include=FALSE}
library(knitr)
opts_chunk$set(warning=FALSE, message=FALSE)

```

## CRAN Task View on Spatial Data

<iframe src="https://cran.r-project.org/web/views/Spatial.html"></iframe>


## Geographic Data Display

* Points, Regions, Routes
* Variables can be continuous and/or discrete data
* Static or Interactive

## Choropleths
```{r}
library(choroplethr)
library(choroplethrMaps)
data(df_state_demographics)
names(df_state_demographics)
```

## Choropleths
```{r}
kable(head(df_state_demographics))
```

## Median Rent: State-level
```{r}

dfstatemedrent=df_state_demographics[,c(1,8)] # Median Rent
colnames(dfstatemedrent)=c("region","value")
state_choropleth(dfstatemedrent, title="Median Rent by State")
```

## Median Rent: County-level

```{r}
data("df_county_demographics")
#names(df_county_demographics)
# Federal Information Processing Standard (FIPS) https://en.wikipedia.org/wiki/FIPS_county_code
kable(head(df_county_demographics))
```

## Median Rent: County-level
```{r}
dfcountymedrent=df_county_demographics[,c(1,8)] 
colnames(dfcountymedrent)=c("region","value")
county_choropleth(dfcountymedrent, title="Median Rent by County")
```

## Population: Specific County/ZCTA
```{r}
library(choroplethrZip)#devtools::install_github("arilamstein/choroplethrZip")
#Zip Code Tabulation Areas are generalized area representations of the United States Postal Service (USPS) ZIP code service areas
library(ggplot2)
data("df_zip_demographics")
kable(head(df_zip_demographics))
```

## Population: Washington

```{r}
dfzipmedrent=df_zip_demographics[,c(1,8)] 
colnames(dfzipmedrent)=c("region","value")
zip_choropleth(dfzipmedrent,state_zoom="washington")+coord_map() #adds mercator projection
```

## Population: Spokane County (FIPS 53063)

```{r}

zip_choropleth(dfzipmedrent, county_zoom=53063) + coord_map()
```

## Interactive map using leaflet, tigris and acs

```{r echo=FALSE}
# Kyle Walker's code has come in very handy
# install and load the packages 
library(leaflet)
library(tigris) #devtools::install_github('walkerke/tigris')
library(acs)
library(stringr)
#lookup_code("Washington","Spokane") #"FIPS code for Washington is '53' and the code for Spokane County is '063'."
spcounty=tracts(state="53",county="063")
# 
# #api.key.install(" YOUR API KEY GOES HERE") 
# 
# geography=geo.make(state="Washington",county="Spokane",tract="*") # from acs package
# rent_data <- acs.fetch(endyear = 2013, 
#                          geography = geography,
#                          variable = "B25064_001")
# 
# 
# 
# rent_df <- data.frame(paste0(as.character(rent_data@geography$state), 
#                                as.character(rent_data@geography$county), 
#                                rent_data@geography$tract), 
#                         rent_data@estimate)
# 
# save(rent_df,file="rent_df.Rda")
load("rent_df.Rda")



colnames(rent_df) <- c("GEOID", "hhrent")

rent_df$GEOID=as.character(rent_df$GEOID)
rent_df$GEOID=str_replace(rent_df$GEOID,"536","5306")

spcounty_merged=geo_join(spcounty, rent_df, "GEOID", "GEOID")


pal <- colorQuantile("Blues", NULL, n = 4)
popup <- paste0("Median gross rent: ", as.character(spcounty_merged$hhrent))

leaflet() %>%
  addProviderTiles("CartoDB.Positron") %>%
  addPolygons(data = spcounty_merged, 
              fillColor = ~pal(spcounty_merged$hhrent), 
              fillOpacity = 0.7, 
              weight = 0.2, 
              popup = popup) %>%
  addLegend(pal = pal, 
            values = spcounty_merged$hhrent, 
            position = "bottomright", 
            title = "Median Gross Rent")
```
  
## Points

```{r}
thingstodo=read.table(text="
                      Attraction  lat lon  Population
                      Jepson  47.667268 -117.405114 45
                      Starbucks  47.669088 -117.396847  50
                      WSU 47.660960 -117.405697 250
                      EWU 47.661061 -117.404044 400",header=TRUE)
```

## Geocoding, if only address is available

```{r}

library(ggmap)

whatislatlon=function(mydata,addressindata){
locs=geocode(as.character(unique(mydata[,addressindata])))
locs$address=unique(mydata[,addressindata])
mydata$latitude=locs$lat[ match(mydata[,addressindata],locs$address)]
mydata$longitude=locs$lon[ match(mydata[,addressindata],locs$address)]
return(mydata)
}
```

## That function works

```{r}

Address=c("502 E Boone Ave, Spokane, WA, 99258","502 E Boone Ave, Spokane, WA, 99258")
mydummydata=data.frame(Address=Address)
mysmartdata=whatislatlon(mydummydata,"Address")
kable(mysmartdata)
```

## A map of location of interest

```{r }
location=c(-117.402209,47.665330)
map=get_map(location=location,maptype="roadmap",source="google",zoom=16)
spokanemap=ggmap(map)
print(spokanemap)
```

## Add Attractions
```{r}
spokanemap=spokanemap+geom_point(data=thingstodo,
            aes(lon,lat,color=Attraction),size=5)
print(spokanemap)
```

## Some Cleaning

```{r}
spokanemap+theme(panel.grid.major = element_blank(),
                 panel.grid.minor = element_blank(),
                axis.text = element_blank(),axis.title = element_blank(),
                axis.ticks = element_blank())
```

## A traveling student's route

```{r}
routes=data.frame(x=thingstodo$lon,y=thingstodo$lat)
newmap=get_googlemap(center=location,zoom=16,
                     markers=routes,
                     path = routes,scale=2,maptype = "satellite")
ggmap(newmap,darken=.3)+geom_text(data=thingstodo,aes(lon,lat,label=Attraction),
                                                  color="white",size=3)
```

## R-Studio's `leaflet` package

* Interface to `leaflet` JS  
```{r}
library(leaflet) #rstudio package
leaflet() %>% addTiles()
```

## Add our points of attraction
```{r}
leaflet() %>% addTiles()%>%addCircleMarkers(data=thingstodo)
```

## Give more information
```{r}
leaflet() %>% addTiles()%>%addCircleMarkers(data=thingstodo,popup=~Attraction,radius=~Population*.05)
```

## Markers
```{r}
leaflet() %>% addTiles()%>%addMarkers(data=thingstodo,popup=~Attraction)
```

## Routes

```{r}
leaflet() %>% addTiles()%>%addMarkers(data=thingstodo,popup=~Attraction)%>% 
  addPolylines(thingstodo$lon,thingstodo$lat)
```

## Few Controls
```{r echo=FALSE}
leaflet() %>% addTiles(group = "OSM (default)") %>%
  addProviderTiles("Stamen.Toner", group = "Toner") %>%
  addProviderTiles("Stamen.TonerLite", group = "Toner Lite") %>%
  addMarkers(data=thingstodo,popup=~Attraction)%>% 
  addPolylines(thingstodo$lon,thingstodo$lat,group="Route")%>%
  addLayersControl(
    baseGroups = c("OSM (default)", "Toner", "Toner Lite"),
    overlayGroups = c("Route"),
    options = layersControlOptions(collapsed = FALSE)
  )%>% hideGroup("Route")
```

## Leaflet+Shiny+DataTable

Some cool stuff possible:[http://147.222.28.7/INRUG-ShinyApp/](http://147.222.28.7/INRUG-ShinyApp/)

## Blog example 1: Air Pollution Levels
[http://patilv.com/airpollution/](http://patilv.com/airpollution/)
<iframe src="http://patilv.github.io/Airpollutionpm/topcitiespm10map.html"></iframe>

## Blog example 2: Mortality Rates of Children under 5 per 1000 live births

[http://patilv.com/MortalityUnder5/](http://patilv.com/MortalityUnder5/)
<iframe src="http://bl.ocks.org/patilv/raw/410a1de459998f35599a/"></iframe>

## Blog example 3: Animated Choropleths

[http://bit.ly/patilanimatedchoropleths](http://bit.ly/patilanimatedchoropleths)

## Blog example 4: Great Circles
[http://patilv.com/USOpenCountries/](http://patilv.com/USOpenCountries/)
<img src="playernationalitymap.png" width=900 height=700>

## Code used is borrowed from many folks including: 

* R-Studio: [https://rstudio.github.io/leaflet/](https://rstudio.github.io/leaflet/)
* Ari Lamstein: [http://www.arilamstein.com/](http://www.arilamstein.com/)
* Kyle Walker: [http://walkerke.github.io/](http://walkerke.github.io/)
* Erik Erhardt: [http://statacumen.com/](http://statacumen.com/)
<h4> Thanks are due to the many package developers and the wonderful R community.