Lesson 5 - ggplot cheatsheet

pacman::p_load(ggplot2, tidyverse, ggthemes, knitr, extrafont, dplyr, scales, lubridate)

setwd("C:/Users/engha/Desktop/Alvin/NYU MSBA/Module 3 - Shanghai/Data Visualization/Lesson 5")
data <- read_csv("bikesharedailydata.csv")

head(data)

summary(data)

data1 <- na.omit(data)
dim(data1)

# customize font
fonts <- theme_bw() + 
  theme(plot.title = element_text(size=14, family="Trebuchet MS", color="#666666", face="bold")) +
  theme(plot.subtitle = element_text(size=10, family="Trebuchet MS", color="#666666")) +
  theme(axis.title = element_text(size=10, family="Trebuchet MS", color="#666666", face="bold")) + 
  theme(axis.title.y = element_text(angle=0)) +
  theme(plot.caption = element_text(size=8, family="Trebuchet MS", color="#666666"))

# remove chart junk
removeJunk <- theme(panel.border = element_blank(),
                    panel.grid = element_blank(),
                    panel.grid.minor = element_blank(),
                    axis.line = element_line(colour = "gray"),
                    axis.ticks.x = element_blank(),
                    axis.ticks.y = element_blank())

# create another column with abbreviated months
data1 <- mutate(data1, month = month.abb[mnth])
data1$month <- factor(data1$month, levels = month.abb)

# pre-attentive attributes
userMnthlyTotal <- data1 %>% 
  group_by(mnth) %>% 
  summarize(total = sum(cnt))
userAverage <- mean(userMnthlyTotal$total)
attributeMean <- geom_hline(yintercept = userAverage, size=1, color="#999999", linetype="dashed")

# plot bar chart
ggplot(data1, aes(x=month, y=cnt)) +
  geom_bar(stat="identity", fill = "seagreen3") + # Basic bar chart
  labs(title="NUMBER OF BIKE RENTALS BY MONTH", # add labels
       subtitle="This chart shows the number of bike rentals by month over the period Jan 2011-Dec 2012. 
The warmer months of May-Oct see an above-average rentals.",
       caption="Source: Alvin Eng, for NYU MSBA Data Visualization by Professor K. Sosulski.",
       x = "Month",
       y="Rentals") +
  fonts + 
  removeJunk +
  scale_y_continuous(labels = comma) + # format scale
  attributeMean +
  annotate("text", x=12, y=300000, label="Average", color="#666666", size=3)

# create column for dates
data1$date <- mdy(data1$dteday)

# plot line chart
ggplot(data1, aes(x=date, y=cnt, group=1)) +
  geom_line(color = "seagreen3") + 
  geom_point(color = "seagreen3", size=1) +
  geom_smooth(method=loess) +
  labs(title="DAILY BIKE RENTALS",
       subtitle="This chart shows the number of daily bike rentals over the period Jan 2011-Dec 2012. 
While daily rentals fluctuate significantly, rental demand has generally trended higher over time.",
       caption="Source: Alvin Eng, for NYU MSBA Data Visualization by Professor K. Sosulski.",
       x = "Date",
       y="Rentals") +
  fonts + 
  removeJunk +
  scale_y_continuous(labels = comma) +
  scale_x_date(date_breaks = "4 month", 
               labels=date_format("%b-%y"),
               limits = as.Date(c('2011-01-01','2012-12-31')))

ggplot(data1, aes(x=date, y=cnt, group=1)) +
  geom_area(fill="seagreen3", color="seagreen1") + 
  geom_smooth(method=loess, color="limegreen", se=FALSE) +
  labs(title="DAILY BIKE RENTALS",
       subtitle="This chart shows the number of daily bike rentals over the period Jan 2011-Dec 2012. 
While daily rentals fluctuate significantly, rental demand has generally trended higher over time.",
       caption="Source: Alvin Eng, for NYU MSBA Data Visualization by Professor K. Sosulski.",
       x = "Date",
       y="Rentals") +
  fonts + 
  removeJunk +
  scale_y_continuous(labels = comma) +
  scale_x_date(date_breaks = "4 month", 
               labels=date_format("%b-%y"),
               limits = as.Date(c('2011-01-01','2012-12-31')))

max = round(max(data1$cnt),-3) # round maximum daily rentals to nearest thousand 

# plot histogram
ggplot(data1, aes(x=cnt)) +
  geom_histogram(breaks=seq(0,max,max/(max/100)), fill="seagreen3") +
  labs(title="HISTOGRAM OF BIKE RENTALS",
       subtitle="This histogram shows the distribution of daily bike rentals over the period Jan 2011-Dec 2012. 
Each bin represents 100 rentals. There was one day with less than 100 rentals.",
       caption="Source: Alvin Eng, for NYU MSBA Data Visualization by Professor K. Sosulski.",
       x = "Rentals",
       y="Frequency") +
  fonts +
  removeJunk +
  scale_x_continuous(labels = comma) 

ggplot(data1, aes(x=cnt)) +
  geom_density(fill="seagreen3") +
  labs(title="DENSITY DISTRIBUTION OF BIKE RENTALS",
       subtitle="This chart shows the density distribution of daily bike rentals over the period Jan 2011-Dec 2012. 
We see a trimodal distribution, with a peak around 4,500 rentals and two lower peaks.",
       caption="Source: Alvin Eng, for NYU MSBA Data Visualization by Professor K. Sosulski.",
       x = "Rentals",
       y="Density") +
  fonts +
  removeJunk +
  scale_x_continuous(labels = comma) + 
  geom_vline(xintercept = 4500, size=1, color="#999999", linetype="dashed")

ggplot(data1, aes(x=month, y=cnt)) +
  geom_boxplot(fill="white", color="seagreen3") +
  labs(title="BOXPLOT OF BIKE RENTALS",
       subtitle="This boxplot shows the key summary statistics of daily bike rentals over the period Jan 2011-Dec 2012. 
The highest median daily rentals was observed in Jul while Mar and Oct saw noticeaably wide ranges.",
       caption="Source: Alvin Eng, for NYU MSBA Data Visualization by Professor K. Sosulski.",
       x = "Month",
       y="Rentals") +
  fonts +
  removeJunk

data1$season[data1$season==1]<- "Winter"
data1$season[data1$season==2]<- "Spring"
data1$season[data1$season==3]<- "Summer"
data1$season[data1$season==4]<- "Fall"
season <- data1$season

# plot scatterplot
ggplot(data1, aes(x=temp, y=cnt)) +
  geom_point(aes(col=season)) +
  geom_smooth(method=lm, color="grey", se=FALSE) +
  labs(title="HOW TEMPERATURE AFFECTS BIKE RENTALS",
       subtitle="This charts plots daily bike rentals against (normalized) temperature over the period Jan 2011-Dec 2012. 
We see a positive relationship; on warmer days rental demand increases.",
       caption="Source: Alvin Eng, for NYU MSBA Data Visualization by Professor K. Sosulski.",
       x = "Temperature (Normalized)",
       y="Rentals") +
  fonts + 
  removeJunk +
  scale_y_continuous(labels = comma)