results.qmd

# Results

```{r}
#| echo: false
#| message: false
#| warning: false
library(tidyverse)
library(arrow)
library(scales)
library(RColorBrewer)
library(patchwork)
library(lubridate)
library(timeDate)
library(ggrepel)

# Use like: scale_y_continuous(labels = millions) + 
millions <- function(x) {
  ifelse(x == 0, "0", label_comma(scale = 1e-6, suffix = "M")(x))
}

# Use like: scale_y_continuous(labels = thousands) + 
thousands <- function(x) {
  ifelse(x == 0, "0", label_comma(scale = 1e-3, suffix = "K")(x))
}

double_scale <- function(x) {
  ifelse(x >= 1000, paste0(x / 1000, "K"), as.character(x))
}

# Use in mutate: mutate(foo= standardize(bar))
standardize <- function(x) {
  (x - mean(x, na.rm = TRUE)) / sd(x, na.rm = TRUE)
}

base_color <- brewer.pal(n = 3, name = "Set2")[1]
secondary_color <- brewer.pal(n=3, name = "Set2")[2]
rain_color <- brewer.pal(n=3, name = "Set2")[3]
classic <- theme_set(theme_minimal())

temp_label <- "Temperature (°C)"

df <- read_parquet("data/complete_weather_and_taxi_data.parquet")
```

```{r}
#| echo: false
# Add a month_abbr column.
df <- df |>
  mutate(month_abbr = factor(month.abb[month], levels = month.abb)) 
```

## A New Normal: How the COVID-19 Pandemic Reshaped Taxi Ridership Patterns

{{< include sections/3/1_dealing_with_the_pandemic.qmd >}}

```{r}
# Filter out pre- and during-Pandemic data.
df <- df |>
  filter(df$date >= as.Date("2021-10-01"))
```
  
## Patterns in Taxi Data 

```{r}
#| message: false
#| warning: false
# Pivot by trip distance bin.
pivot_trip_distance_df <- df |>
  mutate(
    temperature = round(temperature)
  ) |>
  mutate(
    up_to_ten_mile_trips = trip_count - rowSums(across(ends_with("mile_trips")))
  ) |>
  pivot_longer(
    cols=c(ends_with("mile_trips")),
    names_to="trip_type",
    values_to="trip_type_count",
  ) |>
  mutate(trip_type = factor(trip_type, 
    levels = c(
      "half_mile_trips", 
      "one_mile_trips", 
      "two_mile_trips", 
      "three_mile_trips", 
      "five_mile_trips", 
      "up_to_ten_mile_trips"
      ),
    labels = c(
      "Up to Half-Mile Trips",
      "Up to One-Mile Trips",
      "Up to Two-Mile Trips",
      "Up to Three-Mile Trips",
      "Up to Five-Mile Trips",
      "Up to Ten-Mile Trips"
    ),
    ordered = TRUE)) 
```

{{< include sections/3/2_taxi_usage_patterns.qmd >}}

## Patterns in Weather Data

{{< include sections/3/3_patterns_in_weather_data.qmd >}}

## Weather vs. Trips: Can We Define “Nice” Weather from Ridership Patterns? 

{{< include sections/3/4_weather_vs_trips_count.qmd >}}

{{< include sections/3/4_weather_vs_trips.qmd >}}


## Doing the Timeshift

{{< include sections/3/5_timeshifts.qmd >}}