Climate_Data.Rmd

---
title: Working with Climate Data
author:
- ESR 512 - GIS in Geostatistics
- Postgraduate Institute of Science
- University of Peradeniya
- Thusitha Wagalawatta & Daniel Knitter
date: 10/2015
email: daniel.knitter@fu-berlin.de
bibliography: Course_Statistics_SriLanka.bib
csl: harvard1.csl
output:
  html_document:
    toc: false
    theme: united
    pandoc_args: [
      "+RTS", "-K64m",
      "-RTS"
    ]
  pdf_document:
    toc: false
    toc_depth: 2
    number_sections: true
    pandoc_args: [
      "+RTS", "-K64m",
      "-RTS"
    ]
highlight: pygments
---


## Climate data ##
```{r setup, eval=TRUE, echo=FALSE}
setwd("/media/daniel/homebay/Teachings/WS_2015-2016/Statistics_SriLanka/")

##https://ropensci.org/blog/2014/03/13/rnoaa/
## library(rnoaa)
## options(noaakey = "jlufSGYcMMPcKsETXKorwWaIVeaQgwic")

## out <- ncdc(datasetid = "GHCND", stationid = "GHCND:USW00014895", datatypeid = "PRCP",startdate = "2010-05-01", enddate = "2010-10-31")


```

```{r start-climate, eval=TRUE, echo=TRUE}

## FAO
## http://geonetwork3.fao.org/climpag/agroclimdb_en.php

temp <- read.csv("./data/fao_sl_temp.csv", header = TRUE)
temp$DATE <- as.Date(temp$DATE)
#str(temp)
#head(temp)
temp <- temp[complete.cases(temp)==TRUE,]
#str(temp)

library(chron)
temp$YEAR <- years(as.chron(temp$DATE))
temp$MONTH <- months(as.chron(temp$DATE))
temp.y <- split(temp, temp$YEAR)
temp.m <- split(temp, temp$MONTH)
#str(temp.y)
#str(temp.m)

library(plyr)
temp.mean <- lapply(temp.y, FUN = function(x){ddply(x, c("NAME", "LATITUDE", "LONGITUDE"),
                                              summarise,
										      N = length(VALUE),
											  MEAN = mean(VALUE))}
                 )

#str(temp.mean)
#head(temp.mean)

temp.mean <- lapply(temp.mean, FUN = function(x){x[x$N==12,]})
#temp.mean

temp.mean.m <- lapply(temp.m, FUN = function(x){ddply(x, c("NAME", "LATITUDE", "LONGITUDE","YEAR"),
                                                summarise,
												N = length(VALUE),
												MEAN = mean(VALUE))}
                 )

#str(temp.mean.m)
#head(temp.mean.m)

temp.mean.m <- lapply(temp.mean.m, FUN = function(x) {ddply(x, c("NAME", "LATITUDE", "LONGITUDE"),
                                                      summarise, mean_temp = mean(MEAN))})

#str(temp.mean.m)
#head(temp.mean.m)

sl.mean.temp <- ldply(temp.mean.m, data.frame)
#str(sl.mean.temp)
#head(sl.mean.temp)
sl.mean.temp$.id <- as.factor(sl.mean.temp$.id)

ggplot(sl.mean.temp, aes(x=.id, y=mean_temp, group=NAME, col = NAME)) +
        geom_line() +
            geom_point()


prc <- read.csv("./data/fao_sl_prc.csv", header = TRUE)
prc$DATE <- as.Date(prc$DATE)
#head(prc)
prc <- prc[complete.cases(prc)==TRUE,]
#str(prc)

library(chron)
prc$YEAR <- years(as.chron(prc$DATE))
prc$MONTH <- months(as.chron(prc$DATE))
prc.y <- split(prc, prc$YEAR)
prc.m <- split(prc, prc$MONTH)

#str(prc.y)
#str(prc.m)

library(plyr)
prc.sum <- lapply(prc.y, FUN = function(x){ddply(x, c("NAME", "LATITUDE", "LONGITUDE"), summarise,
                 N = length(VALUE),
                 sum = sum(VALUE))}
                 )

#str(prc.sum)
#head(prc.sum)

prc.sum <- lapply(prc.sum, FUN = function(x){x[x$N==12,]})
#prc.sum

prc.sum.m <- lapply(prc.m, FUN = function(x){ddply(x, c("NAME", "LATITUDE", "LONGITUDE","YEAR"),
                                             summarise,
                                             N = length(VALUE),
                                             sum = sum(VALUE))}
                                             )

#str(prc.sum.m)
#head(prc.sum.m)

prc.sum.m <- lapply(prc.sum.m, FUN = function(x) {ddply(x, c("NAME", "LATITUDE", "LONGITUDE"),
                                                  summarise, mean_prc = mean(sum))})

#str(prc.sum.m)
#head(prc.sum.m)

sl.mean.prc <- ldply(prc.sum.m, data.frame)
#str(sl.mean.prc)
#head(sl.mean.prc)
sl.mean.prc$.id <- as.factor(sl.mean.prc$.id)

ggplot(sl.mean.prc, aes(x=.id, y=mean_prc, group=NAME, col = NAME)) +
    geom_line() +
	geom_point()


.

## test wether the places from the two data sets are the same
sl.mean.temp$NAME %in% sl.mean.prc$NAME

cl.stat <- SpatialPointsDataFrame(coords = cbind(sl.mean.temp$LONGITUDE, sl.mean.temp$LATITUDE),
                                  data = data.frame(NAME = sl.mean.temp$NAME),
								  proj4string = CRS("+init=epsg:4326"))
cl.stat <- remove.duplicates(cl.stat)

library(mapdata)

map("worldHires", regions = "Sri Lanka")
points(cl.stat, pch = 19)
map.axes()
title("Cities with climate data available at FAO ")
```