Cyclist at NYC Bridges

In this LBB, i’m going to create visualization to examine the effect of rainfall on the number of cyclists at various bridges in New York. I’ll be working with a dataset of number of cyclist in New York Bridges on April 2016.

1. Loading package

library(ggplot2)
library(GGally)
library(lubridate)
library(reshape2)
library(scales)

2. Import data

bike <- read.csv("nyc-bicycle.csv")
bike

3. Check data dimension

dim(bike)

## [1] 210  11

4. Check data summary

summary(bike)

##        X                           Date                      Day     
##  Min.   :  0.00   2016-04-01 00:00:00:  7   2016-04-01 00:00:00:  7  
##  1st Qu.: 52.25   2016-04-02 00:00:00:  7   2016-04-02 00:00:00:  7  
##  Median :104.50   2016-04-03 00:00:00:  7   2016-04-03 00:00:00:  7  
##  Mean   :104.50   2016-04-04 00:00:00:  7   2016-04-04 00:00:00:  7  
##  3rd Qu.:156.75   2016-04-05 00:00:00:  7   2016-04-05 00:00:00:  7  
##  Max.   :209.00   2016-04-06 00:00:00:  7   2016-04-06 00:00:00:  7  
##                   (Other)            :168   (Other)            :168  
##  High.Temp...F.  Low.Temp...F.   Precipitation Brooklyn.Bridge
##  Min.   :39.90   Min.   :26.10   0      :119   Min.   : 504   
##  1st Qu.:55.00   1st Qu.:44.10   0.01   : 21   1st Qu.:1447   
##  Median :62.10   Median :46.90   0.09   : 21   Median :2380   
##  Mean   :60.58   Mean   :46.41   0.05   :  7   Mean   :2270   
##  3rd Qu.:68.00   3rd Qu.:50.00   0.15   :  7   3rd Qu.:3147   
##  Max.   :81.00   Max.   :66.00   0.16   :  7   Max.   :3871   
##                                  (Other): 28                  
##  Manhattan.Bridge Williamsburg.Bridge Queensboro.Bridge     Total      
##  Min.   : 997     Min.   :1440        Min.   :1306      Min.   : 4335  
##  1st Qu.:2617     1st Qu.:3282        1st Qu.:2457      1st Qu.: 9596  
##  Median :4165     Median :5194        Median :3477      Median :15292  
##  Mean   :4050     Mean   :4862        Mean   :3353      Mean   :14534  
##  3rd Qu.:5309     3rd Qu.:6030        3rd Qu.:4192      3rd Qu.:18315  
##  Max.   :6951     Max.   :7834        Max.   :5032      Max.   :23318  
## 

5. Convert data type

bike$Precipitation <- as.numeric(bike$Precipitation)
bike$Date <- ymd_hms(bike$Date)
bike$Day <- wday(bike$Day, label = T, abbr = F)
bike$Manhattan.Bridge <- as.numeric(bike$Manhattan.Bridge)
bike$Williamsburg.Bridge <- as.numeric(bike$Williamsburg.Bridge)
bike$Queensboro.Bridge <- as.numeric(bike$Queensboro.Bridge)
bike

6. Check correlation

ggcorr(bike,label=T)

## Warning in ggcorr(bike, label = T): data in column(s) 'Date', 'Day' are not
## numeric and were ignored

5. Plot data visualization

Relationship between precipitation levels and total bike crossings

ggplot(bike, aes(x = Precipitation, y = Total)) +
  geom_point() +
  ggtitle("Effect of Precipitation on Number of Cyclists Using New York Bridges ")+
  stat_smooth(method = "lm")+
  labs(y="Cyclist")+
  scale_y_continuous(labels = comma)

Insight: The number of Cyclists tends to decrease as the level of rainfall increases.

Day vs Total Cyclist

library(reshape2)

bridge1 <- melt(bike, id.vars ="Day", measure.vars=c("Brooklyn.Bridge","Manhattan.Bridge","Williamsburg.Bridge","Queensboro.Bridge"))
bridge1

ggplot(bridge1,aes(Day,value))+
  geom_col(aes(fill=variable), position="dodge")+
  scale_y_continuous(label=comma) +
  labs(y="Total Cyclist")

Insight: The above graph shows that everyday the total of cyclist in each NY bridges are almost the same. The mosttotal bicycle user is at Williamsburg bridge.

Precipitation Level vs Cyclist Numbers by Bridge

ggplot(bike, aes(x = Precipitation)) +
  geom_smooth(aes(y = Brooklyn.Bridge, colour = 'Brooklyn.Bridge'), method = 'glm', method.args = list(family = 'poisson')) +
  geom_smooth(aes(y = Manhattan.Bridge, colour = 'Manhattan Bridge'), method = 'glm', method.args = list(family = 'poisson')) +
  geom_smooth(aes(y = Williamsburg.Bridge, colour = 'Williamsburg Bridge'), method = 'glm', method.args = list(family = 'poisson')) +
  geom_smooth(aes(y = Queensboro.Bridge, colour = 'Queensboro Bridge'), method = 'glm', method.args = list(family = 'poisson')) +
  scale_colour_brewer(type = "div", palette = "Dark2") +
  labs(colour = 'Bridge', y='Total Cyclist') +
  ggtitle("Effect of Precipitation Levels on Cyclist Numbers by Bridge")

Insight: This shows the bridges most used by cyclists are Williamsburg Bridge, Manhattan Bridge, Queensboro Bridge and lastly Brooklyn Bridge. They all show the same trend of decreasing usage as rainfall levels increase but Manhattan Bridge shows the strongest trend.