index.Rmd

---
title       : Maps with R
subtitle    : Workshop
author      : Beatriz Martínez
job         : 
framework   : io2012        # {io2012, html5slides, shower, dzslides, ...}
highlighter : highlight.js  # {highlight.js, prettify, highlight}
hitheme     : tomorrow      # 
widgets     : [interactive]            # {mathjax, quiz, bootstrap}
mode        : selfcontained # {standalone, draft}
knit        : slidify::knit2slides
---

## CONTENT

> 1. Read shapefiles into R... (datos geográficos)
> 2. ...and plot some meaningful data. (datos asociados)
> 3. Plot a map on a streetmap (ggplot + ggmap) (referencia)
> 4. Remove the background street map. 
> 5. googleVis


--- 

## 1. Read shapefiles into R... (datos geográficos)

```{r options, include=FALSE}
opts_chunk$set(warning = FALSE, error=FALSE, message=FALSE, tidy=TRUE)

```
#### GET READY
                    
 
 
               
                          
```{r setwd}
wd <- "/Users/martinez/Dropbox/R"

setwd(paste(wd, "/R_maps_workshop_slidify", sep = ""))

# Download the source from http://www.stats.ox.ac.uk/pub/RWin/bin/macosx/mavericks/contrib/3.1/
# install.packages("packages/rgdal_0.8-16.tgz", repos = NULL)
# install.packages("packages/rgeos_0.3-4.tgz", repos = NULL)
library(rgeos)
library(rgdal)

library(sp)
library(maptools)
```

--- 
## 1. Read shapefiles into R... (datos geográficos)
#### POINT SHAPEFILES: the parks
&nbsp;
           
> Load a Shapefile into R  (downladed from [madrid.org](http://www.madrid.org
/nomecalles/DescargaBDTCorte.icm)) 


```{r load_parks, tidy.opts=list(blank=FALSE, width.cutoff=20), }
# get info
getinfo.shape("Shapefiles/MAD_CM_parques_jardines/parques.shp")

# set projection
ED50<-CRS(paste("+proj=utm +zone=30 +ellps=intl +units=m +no_defs"))
# http://cicero.azavea.com/docs/epsg_codes.html equal to 'ED50(ED77) 
# / UTM zone 38N' but Marid is in the UTM zone 30

# read the shapefile
parquesMadrid <- readShapePoints("Shapefiles/MAD_CM_parques_jardines/parques.shp", proj4string = ED50)
```

--- 
## 1. Read shapefiles into R... (datos geográficos)
##### Shapefiles structure
      
      
```{r parks_structure}
str(parquesMadrid, max.level = 2)
parquesMadrid@bbox
````

--- 
## 1. Read shapefiles into R... (datos geográficos)
##### Shapefiles structure
      
```{r parks_structure_2}
head(parquesMadrid@data)
```


--- 
## 1. Read shapefiles into R... (datos geográficos)
##### Simple plot

```{r simple_plot}
# which objtect plot, and by which variable set the color.
plot(parquesMadrid, col = parquesMadrid$MUNICIPIO)

```

--- 
## 1. Read shapefiles into R... (datos geográficos)
##### Simple plot

```{r simple_plot_2}
# plot a subset
parquesMadrid <- parquesMadrid[parquesMadrid$MUNICIPIO == "Madrid", ]
plot(parquesMadrid, col = parquesMadrid$MUNICIPIO)
```

--- 
## 1. Read shapefiles into R... (datos geográficos)
##### Simple plot

```{r simple_plot_3}
# add format attributes
plot(parquesMadrid, pch = 20, cex = 2, col = "springgreen3")
```

--- 
## 1. Read shapefiles into R... (datos geográficos)
#### POLYGON SHAPEFILES: the neighborhoods
&nbsp;
&nbsp;

```{r load_barrios}
getinfo.shape("Shapefiles/MAD_barrios_madrid/barrios_madrid.shp")

barriosMadrid <- readShapePoly("Shapefiles/MAD_barrios_madrid/barrios_madrid.shp", proj4string = ED50)
class(barriosMadrid)
````

--- 
## 1. Read shapefiles into R... (datos geográficos)
#### POLYGON SHAPEFILES: the neighborhoods
```{r barrios_structure}
str(barriosMadrid, max.level = 2)

barriosMadrid@bbox
head(barriosMadrid@data)
```

--- 
## 1. Read shapefiles into R... (datos geográficos)
#### POLYGON SHAPEFILES: the neighborhoods
```{r simple_plot_4}
plot(barriosMadrid)
plot(parquesMadrid, pch = 20, cex = 2, col = "springgreen3", add = TRUE)
```

--- 
## 2. ... and plot some meaningful data (atributos asociados)
&nbsp;&nbsp;
                   
Merge the meaningful data (e.g. neigborhoods density) with the SpatialPolygonDataFrame @data slot.
      
&nbsp; 
         
1. read the neigborhoods density.
2. get ready to merge both data frames 
3. merge them
4. and plot!

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### Read the neigborhoods density. 
downladed from [munimadrid](http://www-2.munimadrid.es/CSE6/control/seleccionDatos?numSerie=14010100010)
&nbsp;
```{r pop_den}
denMad <- read.csv("Data/denBarriosMad.csv", sep = ";", stringsAsFactors = FALSE)
colnames(denMad) <- c("Barrio", "Densidad")
```

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### Get ready both data frames. 
```{r pop_den_2}
head(denMad)
head(barriosMadrid@data)
````


--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### Get ready both data frames: denMad
```{r pop_den_3}

# Get rid off the 'distritos'
# 'barrios' have the structure ##.# Barrioname 
# remove the "."

denMad <- denMad[which(grepl("[0-9]{2}\\.[0-9]", denMad$Barrio) == TRUE), ]
denMad$Barrio <- gsub(".", "", denMad$Barrio, fixed = TRUE)
head(denMad)
````

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### Get ready both data frames: denMad
```{r pop_den_4}
# Detach neighborhood name from neighborhood code
barriosNamesCodes <- do.call(rbind, strsplit(sub(" ", "\\.", denMad$Barrio), "\\."))
head(barriosNamesCodes)
denMad$Barrio <- barriosNamesCodes[ ,2]
denMad$BarrioCod <- barriosNamesCodes[ ,1]
```

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### Get ready both data frames: denMad
```{r pop_den_5}
head(denMad)
head(barriosMadrid@data)
```

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### Get ready both data frames: barriosMadrid SpatialPolygonDataFrame
```{r merge_data}
# Set the proper encoding
barriosMadrid$DESBDT <- iconv(barriosMadrid$DESBDT, "latin1", "UTF-8")

# Detach neighborhood name from neighborhood code
barriosNamesCodes <- strsplit(sub(" ", "\\.", barriosMadrid$DESBDT), "\\.")
barriosNamesCodes <- do.call(rbind, barriosNamesCodes)
head(barriosNamesCodes)
````

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### Get ready both data frames: barriosMadrid SpatialPolygonDataFrame
```{r merge_data_2}
barriosMadrid$BarrioCod <- barriosNamesCodes[ ,1]
barriosMadrid$Barrio <- barriosNamesCodes[ ,2]
head(barriosMadrid@data)
````

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### Merge both data frames
```{r merge_data_3}

length(denMad$Barrio)
length(barriosMadrid$DESBDT)
identical(sort(barriosMadrid$BarrioCod), sort(denMad$BarrioCod))


barriosMadrid@data <- merge(barriosMadrid@data, denMad[ ,c(2,3)])
````

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### Merge both data frames
```{r merge_data_4}

head(barriosMadrid@data)

```

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### and plot!
```{r plot_data}
plot(barriosMadrid, col = barriosMadrid$Densidad)
````

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### and plot!
```{r plot_data_1}
# Plot with SP library using custom colors
library(RColorBrewer)
colors <- colorRampPalette(brewer.pal(9, "YlGnBu"))(16)
spplot(barriosMadrid, "Densidad", col.regions = colors)
```

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### and plot! Customized colors using the `plot` function
```{r plot_data_2}
# Cut the variable into breaks and assign a color to every interval.
cut <- cut(barriosMadrid$Densidad,breaks=c(0,25,50,75,100,150,200,250,300,350,500))
barriosMadrid$cut <- cut
colors <- colorRampPalette(brewer.pal(9, "YlGnBu"))(10)
cutColors <- cut(barriosMadrid$Densidad,breaks=c(0,25,50,75,100,150,200,250,300,350,500), labels = colors)
barriosMadrid$colors <- cutColors

head(barriosMadrid@data)
```

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### and plot! Customized colors using the `plot` function
```{r plot_data_3}
plot(barriosMadrid, col = as.character(barriosMadrid@data$colors))
legend("bottomleft", cex = .7, legend = levels(barriosMadrid$cut), fill = colors, title = "Population density Madrid (2013)")
```

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### and plot! Add neighborhoods labels
```{r labels}
# calculate the polygons centers
# the coordinates function gives the centers of an SpatialPolygonDataFrame
cents <- coordinates(barriosMadrid)
head(cents)
````

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### and plot!
```{r labels_1}
#png("labels.png", width = 2500, height = 2100)
plot(barriosMadrid, col = as.character(barriosMadrid@data$colors))
plot(parquesMadrid, pch = 20, cex = 4, col = "springgreen3", add = TRUE)
pointLabel(cents[ ,1], cents[ ,2], labels=barriosMadrid$Barrio)
legend("bottomleft", cex = .9, legend = levels(barriosMadrid$cut), fill = colors, title = "Population density Madrid (2013)")
#dev.off()
```

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### and plot!: density as points
```{r den_as_points}
# maptools function
# plot as many dots as people living in every polygon
dots <- dotsInPolys(barriosMadrid, barriosMadrid$Densidad)
# png("density.png", width = 2500, height = 2100)
plot(dots, pch = 16, cex = .1, col = "violetred3")
# dev.off()

```

--- 
## 2. ... and plot some meaningful data (atributos asociados)

#### and plot!: density as points over the neighborhood and include the parks
```{r den_as_points_1}
plot(barriosMadrid, col = "grey90")
plot(dots, pch = 16, cex = .1, col = "violetred3", add = TRUE)
plot(parquesMadrid, pch = 20, cex = 1, col = "springgreen3", add = TRUE)
```

--- 
## 2. ... and plot some meaningful data (atributos asociados)
#### and plot!: include the metro lines
```{r metro}
metroLines <- readShapeLines("Shapefiles/MAD_lineas_metro/lineas.shp", proj4string = ED50)

plot(barriosMadrid, col = "grey90")
plot(dots, pch = 16, cex = .1, col = "violetred3", add = TRUE)
plot(parquesMadrid, pch = 20, cex = 1, col = "springgreen3", add = TRUE)
plot(metroLines, col = "turquoise3", lwd = 2, add = TRUE)
```

--- 

## 3. Plot a map on a street map (referencia)

#### ggplot2 + ggmap 
             
&nbsp;
            
The **rgdal** and **rgeos** libraries are required. 
              
&nbsp;
&nbsp;

1. Change the coordinate system, as we have UTM in our objects and ggplot works on **long, lat** coordinates.  
2. Create a dataframe from the Spatial**DataFrame data slot, containing the long and lat variables, as ggplot plots do not plot Spatial objects but dataframes.
3. and plot! (ggplot2 + ggmap)
           


--- 

## 3. Plot a map on a street map (referencia)
#### Change the coordinate system

```{r rgdal}
str(barriosMadrid, max.level = 2)
barriosMadrid@proj4string

# Transform the coordinates to lon lat
barriosLongLat <- spTransform(barriosMadrid, CRS("+proj=longlat"))  
parquesLongLat <- spTransform(parquesMadrid, CRS("+proj=longlat"))
metroLongLat <- spTransform(metroLines, CRS("+proj=longlat"))
```

--- 

## 3. Plot a map on a street map (referencia)
#### Create a dataframe from the Spatial**DataFrame data slot. 
#### Points DataFrame to ggplot

```{r ggplot}
library(ggplot2)

# Points DataFrame to ggplot
head(coordinates(parquesLongLat))

long <- coordinates(parquesLongLat)[, 1]  
lat <- coordinates(parquesLongLat)[, 2]
```

--- 

## 3. Plot a map on a street map (referencia)
#### Create a dataframe from the Spatial**DataFrame data slot. 
#### Points DataFrame to ggplot

```{r ggplot_4}
dataParques <- cbind(parquesLongLat@data, long, lat)
head(dataParques)
```

--- 

## 3. Plot a map on a street map (referencia)
#### Create a dataframe from the Spatial**DataFrame data slot. 
#### Points DataFrame to ggplot

```{r ggplot_5}
ggplot(dataParques, aes(long, lat)) +
geom_point(colour = "springgreen3", size = 4)
```

--- 

## 3. Plot a map on a street map (referencia)
#### Create a dataframe from the Spatial**DataFrame data slot. 
#### Polygons DataFrame to ggplot

```{r ggplot_8}
# The coordinates are the centers
head(coordinates(barriosLongLat))

barriosLongLat@data$id <- rownames(barriosLongLat@data)
head(barriosLongLat@data)
```

--- 

## 3. Plot a map on a street map (referencia)
#### Create a dataframe from the Spatial**DataFrame data slot. 
#### Polygons DataFrame to ggplot

```{r ggplot_6}
# magic
dataBarrios <- fortify(barriosLongLat, region="id") 
class(dataBarrios)
head(dataBarrios)
```

--- 

## 3. Plot a map on a street map (referencia)
#### Create a dataframe from the Spatial**DataFrame data slot. 
#### Polygons DataFrame to ggplot

```{r ggplot_7}
dataBarrios <- merge(dataBarrios, barriosLongLat@data)
head(dataBarrios)
```


--- 

## 3. Plot a map on a street map (referencia)
#### Create a dataframe from the Spatial**DataFrame data slot. 
#### Lines DataFrame to ggplot

```{r ggplot_9}

metroLongLat@data$id <- rownames(metroLongLat@data)
dataMetro <- fortify(metroLongLat, region="id")  # requires rgeos package
dataMetro <- merge(dataMetro, metroLongLat@data)
head(dataMetro)
unique(dataMetro$ETIQUETA)
```

--- 

## 3. Plot a map on a street map (referencia)
#### Plot them! get the map
```{r, warning=FALSE, fig.width=5, fig.height=5}
library(ggmap)
MadridMap <- qmap(location = "Museo del Prado", zoom = 15, source = "google")
MadridMap
```

--- 

## 3. Plot a map on a street map (referencia)
#### Plot them! get the map
```{r warning=FALSE, fig.width=5, fig.height=5}

MadridMap <- qmap(location = barriosLongLat@bbox, zoom = 11, maptype = "terrain")  # Get the map from Google maps
MadridMap
```

--- 

## 3. Plot a map on a street map (referencia)
#### Plot them! get the map
```{r, warning=FALSE, fig.width=5, fig.height=5}

MadridMap + 
  geom_polygon(aes( x = long, y = lat, group = group, fill = Densidad), data = dataBarrios, alpha = 0.5, colour = "white") +
  scale_fill_gradient(low="grey90", high="black", limits=c(0,500)) +
  geom_path(aes(x = long, y = lat, group = group), data = dataMetro, colour = "turquoise3") +
	geom_point(aes(x = long, y = lat), data = dataParques,  , size = 5, alpha = 0.8, colour = "springgreen3")+
	theme(legend.position = "none")
```


--- 
## 4. Plot a map with ggplot without a background street map. 
```{r, warning=FALSE, fig.width=5, fig.height=5}
ggplot() + geom_polygon(aes(x=long, y=lat, group = group, fill = Densidad), data = dataBarrios) + 
	scale_fill_gradient(low="grey90", high="black", limits=c(0,500)) +
	geom_point(aes(x=long, y=lat, size = 5), data = dataParques, colour = "springgreen3") +
	geom_path(aes(x=long, y = lat, group = group), data = dataMetro, colour = "turquoise3") +
	theme(legend.position = "none", 
  	  panel.background = element_blank(),
  	  panel.grid = element_blank(),
  	  axis.ticks = element_blank(),
  	  text = element_blank())
```

--- 
## 5. googleVis
```{r}
library(googleVis)
suppressPackageStartupMessages(library(googleVis))
emigrantes <- read.csv("Data/emigrantes.csv", sep = ";")
head(emigrantes)

# me quedo las que no empiezan por dos mayúsculas
emigrantes <- emigrantes[which(grepl("^[A-Z]{2}", emigrantes$pais) == FALSE), ]
head(emigrantes)
```

--- &interactive
## 5. googleVis
```{r opts.label = 'interactive', results = 'asis'}
emigrantesMap <- gvisGeoChart(emigrantes, locationvar = "pais", colorvar = "X2012", options = list(width = 600, height = 400))
print(emigrantesMap, tag = 'chart')

# plot(emigrantesMap)
# cat(emigrantesMap$html$chart, file="emigrantes2012.html")
```

--- 
                         
&nbsp;
&nbsp;
&nbsp;
&nbsp;
                    
                    
## R to cartoDB
[walkerke RPubs](http://rpubs.com/walkerke/r2cartodb)