EDA_backup.Rmd

---
title: "EDA"
author: "Si Chen"
date: "`r Sys.Date()`"
output: github_document
---

```{r load_library, warning = FALSE, message = FALSE}
library(tidyr)
library(tidyverse)
library(rvest)
library(dplyr)
library(cowplot)
library(gridExtra)
library(RColorBrewer)
library(plotly)

knitr::opts_chunk$set(
    echo = TRUE,
    warning = FALSE,
    fig.width = 8, 
  fig.height = 6,
  out.width = "90%"
)

theme_set(theme_minimal() + theme(legend.position = "bottom"))

options(
  ggplot2.continuous.colour = "viridis",
  ggplot2.continuous.fill = "viridis"
)

scale_colour_discrete = scale_colour_viridis_d
scale_fill_discrete = scale_fill_viridis_d
```

### Load clean datasets

```{r load_data}
df_2013 = 
  read_csv("data/merge_data_2013.csv", show_col_types = FALSE)

df_2017 = 
  read_csv("data/merge_data_2017.csv", show_col_types = FALSE)

# Convert month to factor with levels in ascending order and labels as month abbreviations
df_2013$month <- factor(df_2013$month, levels = 1:12, labels = month.abb[1:12])

df_2013 <- df_2013 %>%
  mutate(date = paste(month, day, sep = "_"))

```

```{r agg_date}
average_delay_by_date <- df_2013 %>%
  # group_by(date) %>%
  group_by(month, date) %>%
  summarise(
    avg_arr_delay = mean(arr_delay, na.rm = TRUE),
    avg_precip = mean(precip, na.rm = TRUE),
    avg_wind_dir = mean(wind_dir, na.rm = TRUE),
    avg_wind_speed = mean(wind_speed, na.rm = TRUE),
    avg_wind_gust = mean(wind_gust, na.rm = TRUE),
    avg_pressure = mean(pressure, na.rm = TRUE),
    avg_visib = mean(visib, na.rm = TRUE)
    )

```



### Data set description

In the dataset for 2013, there are `r ncol(df_2013)` columns and `r nrow(df_2013)` rows. 
In the dataset for 2017, there are `r ncol(df_2017)` columns and `r nrow(df_2017)` rows. 

There are `r n_distinct(df_2013$carrier)` different carriers flying out of NYC airports. NYC has `r n_distinct(df_2013$origin)` different airports. There are `r n_distinct(df_2013$dest)` different destination locations to which flights fly out of NYC airports. 

TBC...


# Weather

Visualize association between weather variables?
aggregate by date, plot scatter plot, add fitted line (instead of by month)?

```{r aggregate_df_2013_month}

# Aggregate the data by month and calculate averages
df_2013_avg <- df_2013 %>%
  group_by(month) %>%
  summarize(
    avg_arr_delay = mean(arr_delay, na.rm = TRUE),
    avg_precip = mean(precip, na.rm = TRUE),
    avg_wind_dir = mean(wind_dir, na.rm = TRUE),
    avg_wind_speed = mean(wind_speed, na.rm = TRUE),
    avg_wind_gust = mean(wind_gust, na.rm = TRUE),
    avg_pressure = mean(pressure, na.rm = TRUE),
    avg_visib = mean(visib, na.rm = TRUE)
  )

```



## Preciptation 2013

```{r precip, echo=FALSE}
# Plot precipitation against month using bar plot
precip_bar =  
  df_2013 %>% 
  ggplot(aes(x = month, y = precip)) +
  geom_bar(stat = "summary", fun = "mean", fill = "skyblue", alpha = 0.7) +
  labs(title = "Average Precipitation by Month",
       x = "Month",
       y = "Precipitation (inches)")

# Plot average precipitation by month in scatter plot + line
precip_line = 
  ggplot(df_2013, aes(x = month, y = precip)) +
  geom_point(stat = "summary", fun = "mean", size = 3, color = "coral") +
  geom_line(aes(group = 1), stat = "summary", fun = "mean", color = "skyblue", size = 1) +
  labs(title = "Average Precipitation by Month",
       x = "Month",
       y = "Precipitation (inches)")

# precipitation by month in violin plot
precip_violin = 
  ggplot(df_2013, aes(x = month, y = precip, color = month)) +
  geom_violin(alpha = 0.7) +
  labs(title = "Distribution of Precipitation by Month",
       x = "Month",
       y = "Precipitation (inches)")


# Plot bar plot with line overlay
precip_line_bar = 
  ggplot(df_2013, aes(x = month, y = precip))  +  
  geom_bar(aes(x=month, y=precip),stat="summary", fun = 'mean', fill="skyblue")+ 
  geom_point(aes(x=month, y=precip), stat = "summary", fun = "mean", size = 2, color = "coral", alpha = .5) + 
  geom_line(aes(group = 1), stat="summary", fun = 'mean', color="coral", size=1, alpha = .8)+ 
  labs(
    title= "Average Precipitation by Month", 
    x="Month",
    y="Precipitation")
  
```


```{r precip_delay}
# Plot precipitation against arrival delay in a scatter plot
precip_delay = 
  ggplot(df_2013, aes(x = precip, y = arr_delay)) +
  geom_point(alpha = 0.5, color = "skyblue") +
  labs(title = "Scatter Plot of Precipitation against Arrival Delay",
       x = "Precipitation (inches)",
       y = "Arrival Delay (minutes)")

```


```{r precip_delay_month}
# Average precip against average arr_delay by month
precip_delay_month = 
  ggplot(df_2013_avg, aes(x = avg_precip, y = avg_arr_delay, color = month)) +
  geom_point(size = 3) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Precipitation against Arrival Delay",
       x = "Precipitation (inches)",
       y = "Arrival Delay (minutes)")

```

```{r precip_delay_date}
precip_delay_date = 
  ggplot(average_delay_by_date, aes(x = avg_precip, y = avg_arr_delay, color = month)) +
  geom_point(size = 3, alpha = .6) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Precipitation against Arrival Delay by date",
       x = "Precipitation (inches)",
       y = "Arrival Delay (minutes)")

precip_delay_date


```



```{r show_precip}

# grid.arrange(precip_line_bar, precip_delay, nrow = 2)
# precip_violin
# precip_line
# precip_bar
summary(df_2013$precip)

precip_line_bar
precip_delay
precip_delay_month
# precip_delay_date

```


## Pressure

```{r plot_pressure}

# Plot average pressure by month in line plot
pressure_line = 
  ggplot(df_2013, aes(x = month, y = pressure))  +  
  #geom_bar(aes(x=month, y=pressure),stat="summary", fun = 'mean', fill="skyblue")+ 
  geom_point(aes(x=month, y=pressure), stat = "summary", fun = "mean", color = "coral", alpha = .8) + 
  geom_line(aes(group = 1), stat="summary", fun = 'mean', color="skyblue", size=1, alpha = .8) + 
  labs(
    title= "Average Pressure by Month", 
    x="Month",
    y="Pressure")

# Plot Pressure against arrival delay in a scatter plot
pressure_delay = 
  ggplot(df_2013, aes(x = pressure, y = arr_delay)) +
  geom_point(alpha = 0.5, color = "skyblue") +
  labs(title = "Scatter Plot of Pressure against Arrival Delay",
       x = "Pressure",
       y = "Arrival Delay (minutes)")

# Average pressure against average arr_delay by month
pressure_delay_month = 
  ggplot(df_2013_avg, aes(x = avg_pressure, y = avg_arr_delay, color = month)) +
  geom_point(size = 3) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Average Pressure against Arrival Delay by Month",
       x = "Avg Monthly Pressure",
       y = "Avg Arrival Delay (minutes)")

```


```{r pressure_date}
# Average pressure against average arr_delay by month
pressure_delay_date = 
  ggplot(average_delay_by_date, aes(x = avg_pressure, y = avg_arr_delay, color = month)) +
  geom_point(size = 3) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Average Pressure against Arrival Delay by Date",
       x = "Avg Pressure by Date",
       y = "Avg Arrival Delay (minutes)")

pressure_delay_date

```



```{r show_pressure}
summary(df_2013$pressure)

pressure_line
pressure_delay
pressure_delay_month

```

## Visibility

```{r visib}

# Plot average Visibility by month in line plot
visib_line = 
  ggplot(df_2013, aes(x = month, y = visib))  +  
  # geom_bar(aes(x=month, y=visib),stat="summary", fun = 'mean', fill="skyblue")+ 
  geom_point(aes(x=month, y=visib), stat = "summary", fun = "mean", color = "coral") + 
  geom_line(aes(group = 1), stat="summary", fun = 'mean', color="skyblue", size=1, alpha = .8) + 
  labs(
    title= "Average Visibility by Month", 
    x="Month",
    y="Visibility")

# Plot Visibility against arrival delay in a scatter plot
visib_delay = 
  ggplot(df_2013, aes(x = visib, y = arr_delay)) +
  geom_point(alpha = 0.5, color = "skyblue") +
  labs(title = "Scatter Plot of Visibility against Arrival Delay",
       x = "Visibility",
       y = "Arrival Delay (minutes)")

# Average Visibility against average arr_delay by month
visib_delay_month = 
  ggplot(df_2013_avg, aes(x = avg_visib, y = avg_arr_delay, color = month)) +
  geom_point(size = 3) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Average Visibility against Arrival Delay by Month",
       x = "Avg Monthly Visibility",
       y = "Avg Arrival Delay (minutes)")


```

```{r visib_date}
# Average Visibility against average arr_delay by date
visib_delay_date = 
  ggplot(average_delay_by_date, aes(x = avg_visib, y = avg_arr_delay, color = month)) +
  geom_point(size = 3) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Average Visibility against Arrival Delay by Date",
       x = "Avg Visibility by Date",
       y = "Avg Arrival Delay (minutes)")


visib_delay_date
```


```{r show_visib}
summary(df_2013$visib)

visib_line
visib_delay
visib_delay_month

```

## Wind Direction

```{r plot_wind_dir}

# Plot average wind_dir by month in line plot
wind_dir_line = 
  ggplot(df_2013, aes(x = month, y = wind_dir))  +  
  geom_bar(aes(x=month, y=wind_dir),stat="summary", fun = 'mean', fill="skyblue", alpha = .8)+ 
  geom_point(aes(x=month, y=wind_dir), stat = "summary", fun = "mean", color = "red") + 
  geom_line(aes(group = 1), stat="summary", fun = 'mean', color="coral", size=1, alpha = .9) + 
  labs(
    title= "Average Wind Direction by Month", 
    x="Month",
    y="Wind Direction")

# Plot Wind Direction against arrival delay in a scatter plot
wind_dir_delay = 
  ggplot(df_2013, aes(x = wind_dir, y = arr_delay)) +
  geom_point(alpha = 0.5, color = "skyblue") +
  labs(title = "Scatter Plot of Wind Direction against Arrival Delay",
       x = "Wind Direction",
       y = "Arrival Delay (minutes)")

# Average Wind Direction against average arr_delay by month
wind_dir_delay_month = 
  ggplot(df_2013_avg, aes(x = avg_wind_dir, y = avg_arr_delay, color = month)) +
  geom_point(size = 3) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Average Wind Direction against Arrival Delay by Month",
       x = "Avg Monthly Wind Direction",
       y = "Avg Arrival Delay (minutes)")

```

```{r wind_dir_date}
# Average Wind Direction against average arr_delay by date
wind_dir_delay_date = 
  ggplot(average_delay_by_date, aes(x = avg_wind_dir, y = avg_arr_delay, color = month)) +
  geom_point(size = 3) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Average Wind Direction against Arrival Delay by Date",
       x = "Avg Daily Wind Direction",
       y = "Avg Arrival Delay (minutes)")

wind_dir_delay_date
```


```{r show_wind_dir}

summary(df_2013$wind_dir)

wind_dir_line
wind_dir_delay
wind_dir_delay_month

```


## Wind Speed
```{r plot_wind_speed}
# Plot average wind_speed by month in line plot
wind_speed_line = 
  ggplot(df_2013, aes(x = month, y = wind_speed))  +  
  geom_bar(aes(x=month, y=wind_speed),stat="summary", fun = 'mean', fill="skyblue", alpha = .8)+ 
  geom_point(aes(x=month, y=wind_speed), stat = "summary", fun = "mean", color = "red") + 
  geom_line(aes(group = 1), stat="summary", fun = 'mean', color="coral", size=1, alpha = .8) + 
  labs(
    title= "Average Wind Speed by Month", 
    x="Month",
    y="Wind Speed")

# Plot wind_speed against arrival delay in a scatter plot
wind_speed_delay = 
  ggplot(df_2013, aes(x = wind_speed, y = arr_delay)) +
  geom_point(alpha = 0.5, color = "skyblue") +
  labs(title = "Scatter Plot of Wind Speed against Arrival Delay",
       x = "Wind Speed",
       y = "Arrival Delay (minutes)")

# Average wind_speed against average arr_delay by month
wind_speed_delay_month = 
  ggplot(df_2013_avg, aes(x = avg_wind_speed, y = avg_arr_delay, color = month)) +
  geom_point(size = 3) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Average Wind Speed against Arrival Delay by Month",
       x = "Avg Monthly Wind Speed",
       y = "Avg Arrival Delay (minutes)")

```


```{r wind_speed_date}
# Average wind_speed against average arr_delay by date
wind_speed_delay_date = 
  ggplot(average_delay_by_date, aes(x = avg_wind_speed, y = avg_arr_delay, color = month)) +
  geom_point(size = 3) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Average Wind Speed against Arrival Delay by Date",
       x = "Avg Daily Wind Speed",
       y = "Avg Arrival Delay (minutes)")

wind_speed_delay_date
```


```{r show_wind_speed}

summary(df_2013$wind_speed)

wind_speed_line
wind_speed_delay
wind_speed_delay_month

```


## Wind Gust
```{r plot_wind_gust}
# Plot average wind_gust by month in line plot
wind_gust_line = 
  ggplot(df_2013, aes(x = month, y = wind_gust))  +  
  geom_bar(aes(x=month, y=wind_gust),stat="summary", fun = 'mean', fill="skyblue", alpha = .8)+ 
  geom_point(aes(x=month, y=wind_gust), stat = "summary", fun = "mean", color = "red") + 
  geom_line(aes(group = 1), stat="summary", fun = 'mean', color="coral", size=1, alpha = .8) + 
  labs(
    title= "Average Wind Gust by Month", 
    x="Month",
    y="Wind Gust")

# Plot wind_gust against arrival delay in a scatter plot
wind_gust_delay = 
  ggplot(df_2013, aes(x = wind_gust, y = arr_delay)) +
  geom_point(alpha = 0.5, color = "skyblue") +
  labs(title = "Scatter Plot of Wind Gust against Arrival Delay",
       x = "Wind Gust",
       y = "Arrival Delay (minutes)")

# Average wind_gust against average arr_delay by month
wind_gust_delay_month = 
  ggplot(df_2013_avg, aes(x = avg_wind_gust, y = avg_arr_delay, color = month)) +
  geom_point(size = 3) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Average Wind Gust against Arrival Delay by Month",
       x = "Avg Monthly Wind Gust",
       y = "Avg Arrival Delay (minutes)")


```

```{r wind_gust_date}
# Average wind_gust against average arr_delay by date
wind_gust_delay_date = 
  ggplot(average_delay_by_date, aes(x = avg_wind_gust, y = avg_arr_delay, color = month)) +
  geom_point(size = 3) +
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) +
  labs(title = "Scatter Plot of Average Wind Gust against Arrival Delay by Date",
       x = "Avg Daily Wind Gust",
       y = "Avg Arrival Delay (minutes)")

wind_gust_delay_date
```


```{r show_wind_gust}

summary(df_2013$wind_gust)

wind_gust_line
wind_gust_delay
wind_gust_delay_month

```


## Weather Variables
```{r plot_weather}

combined_weather_1 <- 
  grid.arrange(
    pressure_delay_date, 
    wind_dir_delay_date, 
    wind_gust_delay_date, 
    wind_speed_delay_date, 
    ncol = 2,
    layout_matrix = matrix(c(1, 2, 3, 4), ncol = 2, byrow = TRUE),
    heights = c(1, 1), widths = c(1, 1),
    top = "Weather Variables")

combined_weather_1

```


# Flights

## Arrival Delay Summary

```{r delay_summary}

summary(df_2013$arr_delay)

dd_hist = 
  ggplot(data = df_2013, aes(x = arr_delay)) +
  geom_histogram(fill = 'skyblue', color = 'coral')+
  labs(title = "Histogram of Arrival Delay",
       x = "Arrival Delay (minutes)",
       y = "Frequency")

dd_hist
```

## Distance - Arrival Delay - Destination

The relationship between the distance and the average delay for each location.

```{r dist_delay_dest}

df_2013_dest <- 
  df_2013 %>% 
  group_by(dest) %>%
  summarize(
    count = n(),
    dist = mean(distance, na.rm = TRUE),
    delay = mean(arr_delay, na.rm = TRUE)
  )

```


```{r}
df_2013_dest

ggplot(df_2013_dest, aes(dist, delay)) + 
  geom_point(aes(size = count, color = dist),alpha = 0.5) + 
  geom_smooth(color = 'skyblue') + 
  geom_jitter() +
  scale_fill_brewer(palette = "Set3") +
  labs(title = "Distance against Arrival Delay",
       x = "Distance",
       y = "Arrival Delay (minutes)")

```

Flights tend to have more delays in short-medium distances. Long distance flights do not have as much delay.



## Arrival delay by origins
```{r origin_delay}

df_2013 %>%
  group_by(origin) %>%
  summarise(mean_ad = mean(arr_delay), sd_ad = sd(arr_delay), IQR_ad = IQR(arr_delay), n = n()) %>% 
  arrange(desc(mean_ad))

# Plot violin plot of arr_delay against origin, coloring by origin
# reversing the order from high to low
delay_origin_box = 
  ggplot(df_2013, aes(x = reorder(origin, -arr_delay), y = arr_delay, fill = origin)) +
  geom_boxplot(trim = FALSE, scale = "width", alpha = 0.7) +
  labs(title = "Violin Plot of Arrival Delay by Origin",
       x = "Origin",
       y = "Arrival Delay (minutes)")

delay_origin_box

```


## Arrival delay by Destination

```{r dest_delay}

df_2013 %>%
  group_by(dest) %>%
  summarise(mean_ad = mean(arr_delay), sd_ad = sd(arr_delay), IQR_ad = IQR(arr_delay), n = n()) %>% 
  arrange(desc(mean_ad))

# Create a summary table with average arrival delay for each destination
avg_arr_delay <- df_2013 %>%
  group_by(dest) %>%
  summarise(avg_arr_delay = mean(arr_delay, na.rm = TRUE))


```

```{r}
# Create an interactive map using plotly
fig <- plot_ly(
  data = avg_arr_delay,
  type = 'scattergeo',
  locationmode = 'USA-states',
  locations = ~dest,
  z = ~avg_arr_delay,
  text = ~paste('Destination: ', dest, '<br>Avg Arrival Delay: ', round(avg_arr_delay, 2)),
  color = ~avg_arr_delay,
  colorscale = 'Viridis',
  marker = list(line = list(width = 0.5, color = 'white'))
) %>%
  layout(
    geo = list(
      scope = 'usa',
      projection = list(type = 'albers usa'),
      showland = TRUE,
      landcolor = 'rgb(217, 217, 217)',
      subunitwidth = 1,
      countrywidth = 1,
      subunitcolor = 'rgb(255,255,255)',
      countrycolor = 'rgb(255,255,255)'
    ),
    title = 'Average Arrival Delay by Destination',
    xaxis = list(title = 'Destination'),
    yaxis = list(title = 'Average Arrival Delay')
  )

# Show the interactive map
fig


```


```{r plot_dest_delay}
# Calculate average arrival delay by destination
avg_arr_delay_by_dest <- df_2013 %>%
  group_by(dest) %>%
  summarize(avg_arr_delay_dest = mean(arr_delay, na.rm = TRUE)) %>%
  arrange(desc(avg_arr_delay_dest)) %>%
  head(10)

# Select only the top 6 destinations
top_destinations <- avg_arr_delay_by_dest$dest

# Filter the data for the top 6 destinations
filtered_data <- df_2013 %>% filter(dest %in% top_destinations)

# Create a boxplot for the top 6 destinations
top10_dest = 
  ggplot(filtered_data, aes(x = reorder(dest, -arr_delay), y = arr_delay, fill = dest)) +
  geom_boxplot() +
  scale_fill_brewer(palette = "Set3") +
  labs(title = "Boxplot of Arrival Delay for Top 6 Destinations",
       x = "Destination",
       y = "Arrival Delay (minutes)")

 top10_dest
```


## Arrival delay by carriers

```{r carrier_delay}

df_2013 %>%
  group_by(carrier) %>%
  summarise(mean_ad = mean(arr_delay), sd_ad = sd(arr_delay), IQR_ad = IQR(arr_delay), n = n()) %>% 
  arrange(desc(mean_ad))

# Plot violin plot of arr_delay against carrier, coloring by carrier
# reversing the order from high to low
delay_carrier_box = 
  ggplot(df_2013, aes(x = reorder(carrier, -arr_delay), y = arr_delay, fill = carrier)) +
  geom_violin(trim = FALSE, scale = "width", alpha = 0.7) +
  labs(title = "Violin Plot of Arrival Delay by Carrier",
       x = "Carrier",
       y = "Arrival Delay (minutes)")

delay_carrier_box
```


## Arrival Delay by Month

```{r month_delay}

df_2013 %>%
  group_by(month) %>%
  summarise(mean_ad = mean(arr_delay), sd_ad = sd(arr_delay), IQR_ad = IQR(arr_delay), n = n()) %>% 
  arrange(desc(mean_ad))

# Plot violin plot of arr_delay against month, coloring by month
# reversing the order from high to low
delay_month_box = 
  ggplot(df_2013, aes(x = reorder(month, -arr_delay), y = arr_delay, fill = month)) +
  geom_violin(trim = FALSE, scale = "width", alpha = 0.7) +
  labs(title = "Violin Plot of Arrival Delay by Month",
       x = "Month",
       y = "Arrival Delay (minutes)")

delay_month_box
```


## Flight 


## Tailnum

```{r flight_delay}

flights_delay <- 
  df_2013 %>% 
  group_by(tailnum) %>%
  summarize(
    count = n(),
    avg_delay = mean(arr_delay)
  )

```


```{r}
flights_delay

ggplot(flights_delay, aes(avg_delay)) + 
  geom_freqpoly(color = 'skyblue') +
  labs(title = "Distribution of Average Arrival Delay by Flight",
       x = "Average Arrival Delay"
       )

```

```{r find_max}

max_delay <- 
  flights_delay %>% 
  arrange(-avg_delay)

max_delay

```

N305AS had an average delay of 196 minutes, which is about 3 hours. 



## On time arrival rate for NYC airports

Suppose a flight that is delayed for less than 5 minutes is basically “on time”. Assume we onsider any flight delayed for 5 minutes of more to be “delayed”

```{r on_time_arr}
df_2013 <- 
  df_2013 %>% 
  mutate(
    ontime = arr_delay < 5
    )

df_2013 <- 
  df_2013 %>%
  mutate(
    arr_type = ifelse(arr_delay < 5, "on time", "delayed")
    )

df_2013 %>% 
  group_by(origin) %>% 
  summarise(ontime_prop = sum(ontime == TRUE) / n()) %>% 
  arrange(desc(ontime_prop))

ggplot(data = df_2013, aes(x = origin, fill = arr_type)) +
  geom_bar()+
  scale_fill_brewer(palette = "Set3") + 
  labs(title = "Bar Plot of On-time Arrival for NYC Airports",
       x = "Origin",
       y = "Count of on-time arrival")
```

The plot shows that LGA airport has the best on time arrival rate and EWR airport has the worst.



## Average Speed for flight (air_time & distance)

We set average speed = distance / (air_time/60) in miles/hr.

```{r avg_speed}
df_2013 <- 
  df_2013 %>% 
  mutate(
    avg_speed = distance / (air_time/60)
    )

ggplot(data = df_2013, aes(x = distance, y = avg_speed, color = avg_speed)) + 
  geom_point(alpha = .6) +
  scale_fill_brewer(palette = "Set3") + 
  labs(title = "Scatter Plot of Average Flight Speed against Distance",
       x = "Distance",
       y = "Average Flight Speed")

ggplot(data = df_2013, aes(x = avg_speed, y = arr_delay, color = avg_speed)) + 
  geom_point(alpha = .4) +
  scale_fill_brewer(palette = "Set3") + 
  geom_smooth(method = "lm", se = FALSE, color = "coral", size = 0.6) + 
  labs(title = "Scatter Plot of Average Flight Speed against Arrival Delay",
       x = "Average Flight Speed",
       y = "Arrival Delay (minutes)")

```