project_temp_100.Rmd

---
title: "project_temp_100"
author: "JosephKBS"
date: "10/14/2020"
output: html_document
---

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


1. Fetching datasets from the files
```{r}
library(ggplot2)
library(ggthemes)
co2emi <- read.csv("./data/annual-co2-emissions-per-country (1).csv")[22851:23019,3:4]
temp_ano <-  read.csv("./data/temperature-anomaly.csv")[1:169,2:3]
 glotemp <-  read.csv("./data/globaltempfrom1880to2019.csv", header = TRUE)[2:141,1:14]
```

2. data cleaning 
```{r}
# change to dataframe
df <- data.frame( rbind(co2emi[,2], temp_ano[,2], temp_ano[,1] ))
# transpose
df <- as.data.frame(t(as.matrix(df)))
# rowname 
rownames(df)<-c()
colnames(df) <- c("co2", "temp_anomaly", "year")
head(df)
```

3. checking the types of data
```{r}
str(df)
```

4. scaling the data (difference in unit of measurement)
```{r}
df_norm <- scale(df[,1:2])
df_norm <- data.frame(  cbind( df_norm, df$year) )
colnames(df_norm) <- c("co2" ,"temp_anomaly", "year")
head(df_norm)
```
5. visualize the data
```{r}
ggplot(df_norm, aes(year))+
  geom_line(aes(y=co2, col="co2 emission") ) + 
  geom_line(aes(y=df_norm$temp_anomaly, col="temp anomaly")) +
  labs(title="Global temperature anomaly and co2 emission", 
    caption="Source: U.S. EPICA, NOAA", x="Years", y="Scaled carbon dioxide emission")+
    theme_economist() 
```

6. Looking at the new data 
```{r}
ggplot(df_norm, aes(x=year, y=temp_anomaly)) + 
        geom_point( color="blue", alpha=0.3, size=2) +
        geom_line(aes(y=co2, col="co2")) +
  labs(title="Carbon Dioxide Emission and global temperature anomlay", 
    subtitle="World wide data", 
    caption="Source: U.S. EPICA",x="Years", y="Amount(scaled)")+
    theme_classic() 
```
To check whether the type of the data.
```{r}
 df_g <- data.frame( glotemp )
str(df_g)
```

adjust the length of data
```{r}
 df_g <- data.frame( glotemp )

df_g_y<- as.numeric(df_g[,1])
df_g_1<- as.numeric(df_g[,2])
df_g_2<- as.numeric(df_g[,3])
df_g_3<- as.numeric(df_g[,4])
df_g_4<- as.numeric(df_g[,5])
df_g_5<- as.numeric(df_g[,6])
df_g_6<- as.numeric(df_g[,7])
df_g_7<- as.numeric(df_g[,8])
df_g_8<- as.numeric(df_g[,9])
df_g_9<- as.numeric(df_g[,10])
df_g_10<- as.numeric(df_g[,11])
df_g_11<- as.numeric(df_g[,12])
df_g_12<- as.numeric(df_g[,13])
df_g_a<- as.numeric(df_g[,14])
```

creating dataset
```{r}
df_g <- data.frame( rbind(df_g_y, df_g_1, df_g_2, df_g_3, df_g_4, df_g_5, df_g_6, df_g_7, df_g_8, df_g_9,df_g_10,df_g_11, df_g_12, df_g_a) )

df_g <- as.data.frame(t(as.matrix(df_g)))
rownames(df_g)<-c()
colnames(df_g) <- c("year","jan","feb","mar","apr","may","june","july","aug","sep","oct","nov","dec","ave")
df_g <- df_g[1:139,]
head(df_g)
str(df_g)
```

creating a new dataframe 
```{r}
new_data <- data.frame(matrix(NA, nrow=12*139, ncol=3 ))

for (i in 1:139){
  for(j in 1:12){
    
  new_data[i+(139)*(j-1), 1] <- 1879+i
  
  new_data[i+(139)*(j-1), 2] <- df_g[i,j+1]
  new_data[i+(139)*(11), 2] <- df_g[i, 12]
  
  new_data[i+(139)*0,3] <- "Jan"
  new_data[i+(139)*1,3] <- "Feb"
  new_data[i+(139)*2,3] <- "Mar"
  new_data[i+(139)*3,3] <- "Apr"
  new_data[i+(139)*4,3] <- "May"
  new_data[i+(139)*5,3] <- "Jun"
  new_data[i+(139)*6,3] <- "Jul"
  new_data[i+(139)*7,3] <- "Aug"
  new_data[i+(139)*8,3] <- "Sep"
  new_data[i+(139)*9,3] <- "Oct"
  new_data[i+(139)*10,3] <- "Nov"
  new_data[i+(139)*11,3] <- "Dec"  
  }
}

ridge <-  new_data
ridge <- ridge[,1:3]
colnames(ridge) <- c("Year","Temp","Month")

ridge$year <- as.integer(ridge$Year)
ridge$temp <- as.numeric(ridge$Temp)
ridge$month <- as.factor(ridge$Month)

head(ridge)
str(ridge)
```
Creating a ridge plot
```{r}
library(ggrepel)
library(grid)
library(ggridges)

p1 <- ggplot(ridge[,1:3], aes(x=Temp, y = as.factor(Year), 
                        fill = stat(x))) + 
  geom_density_ridges_gradient(scale = 40, rel_min_height = 0.01) +
  scale_fill_viridis_c(name = "Mean.[F]", option = "A") +
  xlim(-0.8, 1.5) +
  theme_classic() +
  labs(title = 'Increasing Mean Temperature', 
       subtitle="Changes in Global Monthly Mean Temperature",
    #   caption="Source: NOAA National Climate Research Unit"
       x="Mean Temperatures", y=" Years (1880-2018) ") + 
        theme(
        #axis.text.x = element_blank(),
        #axis.ticks.x = element_blank(),
        axis.text.y = element_blank(),
        axis.ticks.y = element_blank()) + 
        annotate(geom="text", x=1.3, y=10, label="1880",
              color="black", size=5) +
        annotate(geom="text", x=1.3, y=160, label="2018",
              color="black", size=5)

p1
```

scale the co2 emission data
```{r}
co2emi <- read.csv("./data/annual-co2-emissions-per-country (1).csv")[22851:23019,3:4]

dim(co2emi)
co2emi <- co2emi[31:169,] # from 1880 to 2018
head( co2emi )

#scaling co2 emission data
co2_scale <- scale(co2emi[,2])
co2 <- data.frame(  cbind( co2emi, co2_scale) )
colnames(co2) <- c("Year" ,"co2", "co2_scale")
head(co2)

co2_table <- data.frame(matrix(NA, nrow=12*139, ncol=3 ))

for ( i in 1:139){
  for(j in 1:12){
  co2_table[i+(139)*(j-1), 1] <- co2[i,1]
  co2_table[i+(139)*(j-1), 2] <- co2[i,2]
  co2_table[i+(139)*(j-1), 3] <- co2[i,3]
  }
}
head(co2_table)
colnames(co2_table) <- c("Year" ,"co2", "co2_scale")

# ridge dataset scaling (monthly temp)
ridge <- ridge[,1:3]
ridge_scale <- scale(ridge$Temp)
ridge_data <- data.frame(  cbind( ridge, ridge_scale) )
colnames(ridge_data) <- c("Year","Temp","Month","Scale")
head(ridge_data)

```

# ordering month from Jan to Dec
```{r}
unq <- unique(ridge_data$Month)
ridge_data$Month <- factor(unq, levels = month.abb)

```


# plotting monthly temperature anomaly compared to co2 emission.(scaled)
```{r}

p <- ggplot(ridge_data, aes(x=Year, y=Scale, color=factor(Month))) + 
        geom_point() + 
        scale_colour_tableau(palette="Classic Purple-Gray 12") +
       # scale_fill_discrete( labels = c("Month")) +
    labs(title="Increasing Carbon Dioxide emission and global temperature", 
    subtitle="Total amount of carbon dioxide emission and monthly temperature anomaly compared to 1950-1980.", 
    x="Years", 
    y="Temperature anomaly & co2 emission(Metric Ton, scaled)" ) + theme_classic()       

p2 <- p + geom_line(aes(y=co2_table$co2_scale, col="Co2_Emission"), 
                    size=1.5, color="#990000", label="co2") +  
  annotate(
    geom = "curve", x = 1970, y = 2, 
    xend = 1976, yend = 0.7, 
    curvature = .3, 
    arrow = arrow(length = unit(2, "mm")), 
    color="red", size=1.1
  ) +
  annotate(geom = "text", 
           x = 1950, 
           y = 2.3, 
           label = "Total Co2 Emission", 
           hjust = "left", 
           color="red")
p2
```

# adding up two plots
 description :  https://cran.r-project.org/web/packages/ggpubr/ggpubr.pdf

```{r}
library(ggpubr)
ggarrange(p1, p2, widths = c(1.5,2))
```