Kaggle: S6 E5 Playground

Predicting F1 Pit Stops

train <- read.csv("train.csv")
test <- read.csv("test.csv")

Missing Values

In the given datasets, we have no missing values.

missing_train <- train |>
  dplyr::mutate(across(where(is.character), ~ dplyr::na_if(.x, ""))) |>
  dplyr::summarise(across(everything(), ~ sum(is.na(.x)))) |>
  tidyr::pivot_longer(cols = everything(), names_to = "Variables", values_to = "Total_train")

missing_test <- test |>
  dplyr::mutate(across(where(is.character), ~ dplyr::na_if(.x, ""))) |>
  dplyr::summarise(across(everything(), ~ sum(is.na(.x)))) |>
  tidyr::pivot_longer(cols = everything(), names_to = "Variables", values_to = "Total_test")

missing <- dplyr::left_join(missing_train, missing_test, by="Variables")

missing |> gt()

Variables	Total_train	Total_test
id	0	0
Driver	0	0
Compound	0	0
Race	0	0
Year	0	0
PitStop	0	0
LapNumber	0	0
Stint	0	0
TyreLife	0	0
Position	0	0
LapTime..s.	0	0
LapTime_Delta	0	0
Cumulative_Degradation	0	0
RaceProgress	0	0
Position_Change	0	0
PitNextLap	0	NA

Data Exploration

Class Imbalance Check

target_imbalance <- as.data.frame(table(train$PitNextLap)) 
colnames(target_imbalance) <- c("Target", "Frquency")
target_imbalance |>  
  gt()

Target	Frquency
0	351759
1	87381

Data Correlations

train_numeric <- train |>
  mutate(
    Driver = as.numeric(as.factor(Driver)), 
    Compound = as.numeric(as.factor(Compound)),
    Race = as.numeric(as.factor(Race))
  )

pit_corr <- stats::cor(train_numeric, use = "complete.obs")

corrplot(pit_corr, method = 'color', order = 'alphabet', , addCoef.col = 'black', col = COL2('RdYlBu'))

p1 <- ggplot(train, aes(x = Cumulative_Degradation, fill = factor(PitNextLap))) +
  geom_density(alpha = 0.5) +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Cumulative Degradation",
    y = "Density",
    x = "Cumulative Degradation"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

p2 <- ggplot(train, aes(x = LapNumber, fill = factor(PitNextLap))) +
  geom_density(alpha = 0.5) +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Lap Number",
    y = "Density",
    x = "Lap Number"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

p3 <- ggplot(train, aes(x = Stint, fill = factor(PitNextLap))) +
  geom_density(alpha = 0.5) +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Stint",
    y = "Density",
    x = "Stint"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

p4 <- ggplot(train, aes(x = TyreLife, fill = factor(PitNextLap))) +
  geom_density(alpha = 0.5) +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Tyre Life",
    y = "Density",
    x = "Tyre Life"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

p5 <- ggplot(train, aes(x = LapTime..s., fill = factor(PitNextLap))) +
  geom_density(alpha = 0.5) +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Lap Time",
    y = "Density",
    x = "Lap Time"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

p6 <- ggplot(train, aes(x = LapTime_Delta, fill = factor(PitNextLap))) +
  geom_density(alpha = 0.5) +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Lap Time Delta",
    y = "Density",
    x = "Lap Time Delta"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

p7 <- ggplot(train, aes(x = Position, fill = factor(PitNextLap))) +
  geom_density(alpha = 0.5) +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Position",
    y = "Density",
    x = "Position"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

p8 <- ggplot(train, aes(x = Position_Change, fill = factor(PitNextLap))) +
  geom_density(alpha = 0.5) +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Position Change",
    y = "Density",
    x = "Position Change"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

p9 <- ggplot(train, aes(x = RaceProgress, fill = factor(PitNextLap))) +
  geom_density(alpha = 0.5) +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Race Progress",
    y = "Density",
    x = "Race Progress"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

grid.arrange(p1, p2, p3, p4, p5, p6, p7, p8, p9, ncol=3, nrow=3)

Tyre Life and Degradation

ggplot(train, aes(x = factor(PitNextLap), y = TyreLife, fill = factor(PitNextLap))) +
  geom_boxplot() +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Tyre Life at Time of Pit Decision",
    y = "Tyre Life",
    x = "Pit Next Lap"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

ggplot(train, aes(x = TyreLife, y = Cumulative_Degradation, color = Compound)) +
  geom_smooth() +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Degradation Curve by Compound",
    y = "Cumulative Degradation",
    x = "Tyre Life"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

ggplot(train, aes(x = TyreLife, y = Compound, color = factor(PitNextLap))) +
  geom_boxplot() +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Degradation Curve by Compound",
    y = "Cumulative Degradation",
    x = "Tyre Life"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

Pit Rate by Compound

train |>
  group_by(Compound) |>
  summarise(pit_rate = mean(PitNextLap), count = n()) |>
  ggplot(aes(x = reorder(Compound, pit_rate), y = pit_rate, fill = Compound)) +
  geom_col() +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Pit Rate by Compound",
    y = "Pit Rate",
    x = "Compound"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

Race Progress

ggplot(train, aes(x = RaceProgress, fill = factor(PitNextLap))) +
  geom_density(alpha = 0.5) +
  facet_wrap(~Compound) +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Race Progress Distribution by Pit Decision",
    y = "Density",
    x = "Race Progress"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

Driver Position

ggplot(train, aes(x = Position, fill = factor(PitNextLap), color = factor(PitNextLap))) +
  geom_histogram(position = "fill", bins = 20, alpha=0.3) +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Pit Rate by Track Position",
    y = "Proportion",
   # x = "Race Progress"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

train |>
  group_by(Stint, Compound) |>
  summarise(pit_rate = mean(PitNextLap)) |>
  ggplot(aes(x = Stint, y = pit_rate, color = Compound, fill = Compound)) +
  geom_line() + 
  geom_point() +
  scale_fill_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  scale_color_manual(values = c("#E69F00", "#56B4E9", "#009E73", "#F0E442", 
                                "#0072B2", "#D55E00", "#CC79A7")) +
  labs(
    title = "Pit Rate by Stint Number",
    y = "Pit Rate",
    x = "Stint"
  ) +
  theme_minimal(base_size = 12, base_family = "sans") +
  theme(
    plot.title = element_text(face = "bold", size = 14, margin = margin(b = 4)),
    plot.subtitle = element_text(color = "grey40", size = 11, margin = margin(b = 12)),
    plot.caption = element_text(color = "grey60", size = 9, hjust = 0),
    plot.margin = margin(16, 16, 16, 16),
    panel.grid.major.x = element_blank(),
    panel.grid.minor = element_blank(),
    axis.title = element_text(color = "grey30", size = 10),
    axis.text = element_text(color = "grey30"),
    axis.ticks.x = element_line(color = "grey80"),
    legend.position = "bottom",
    legend.title = element_blank()
  )

race_stint <- train |> 
  group_by(Race, Compound) |>
  summarise(
    min_val  = min(Stint, na.rm = TRUE),
    mean_val = mean(Stint, na.rm = TRUE),
    max_val  = max(Stint, na.rm = TRUE)
  ) |>
  ungroup() 

race_stint |> 
  gt(
    rowname_col = "Compound",
    groupname_col = "Race"
  ) |>
  tab_header("Race Stint Analysis") |>
  cols_width(
    everything() ~ px(150)
    )

	min_val	mean_val	max_val
Race Stint Analysis
Abu Dhabi Grand Prix
HARD	1	2.303222	5
INTERMEDIATE	4	4.000000	4
MEDIUM	1	1.119048	5
SOFT	1	3.800000	5
Australian Grand Prix
HARD	1	2.434992	6
INTERMEDIATE	1	3.974941	7
MEDIUM	1	1.101116	6
SOFT	1	2.700000	6
Austrian Grand Prix
HARD	1	2.603384	7
MEDIUM	1	1.532366	6
SOFT	1	3.160714	7
Azerbaijan Grand Prix
HARD	1	1.896008	5
INTERMEDIATE	2	2.500000	3
MEDIUM	1	1.038069	4
SOFT	3	3.625000	4
Bahrain Grand Prix
HARD	1	2.590173	6
INTERMEDIATE	6	6.000000	6
MEDIUM	1	2.178304	5
SOFT	1	1.607715	6
Belgian Grand Prix
HARD	1	2.433742	4
INTERMEDIATE	1	1.016854	4
MEDIUM	1	1.596089	4
SOFT	1	2.511943	6
British Grand Prix
HARD	1	2.563769	5
INTERMEDIATE	1	1.833844	6
MEDIUM	1	1.287798	6
SOFT	1	3.029770	6
Canadian Grand Prix
HARD	1	2.043497	6
INTERMEDIATE	1	1.165700	6
MEDIUM	1	1.460700	6
SOFT	1	4.305556	6
Chinese Grand Prix
HARD	1	2.330100	5
MEDIUM	1	1.142687	5
SOFT	1	3.000000	5
Dutch Grand Prix
HARD	1	2.253925	7
INTERMEDIATE	1	4.672496	8
MEDIUM	1	1.213644	8
SOFT	1	2.801603	8
Emilia Romagna Grand Prix
HARD	1	2.208805	5
INTERMEDIATE	1	1.010131	4
MEDIUM	1	1.413910	5
SOFT	3	3.602564	5
French Grand Prix
HARD	1	2.169747	4
MEDIUM	1	1.026102	4
Hungarian Grand Prix
HARD	1	2.412507	5
MEDIUM	1	1.285714	6
SOFT	1	2.264993	5
Italian Grand Prix
HARD	1	2.055583	4
INTERMEDIATE	1	1.000000	1
MEDIUM	1	1.121868	4
SOFT	1	2.650000	4
Japanese Grand Prix
HARD	1	2.700276	6
INTERMEDIATE	1	2.854722	5
MEDIUM	1	1.195066	5
SOFT	1	2.867769	5
WET	1	2.037037	4
Las Vegas Grand Prix
HARD	1	2.338379	5
MEDIUM	1	1.045455	4
SOFT	1	3.736842	5
Mexico City Grand Prix
HARD	1	2.169914	6
MEDIUM	1	1.189163	5
SOFT	1	1.939289	5
Miami Grand Prix
HARD	1	1.729888	6
MEDIUM	1	1.244020	5
SOFT	1	3.012500	4
Monaco Grand Prix
HARD	1	1.894737	6
INTERMEDIATE	1	3.068104	8
MEDIUM	1	1.716032	5
SOFT	3	3.965000	6
WET	1	1.136398	6
Pre-Season Testing
HARD	1	2.248302	5
INTERMEDIATE	1	4.000000	6
MEDIUM	1	1.158844	5
SOFT	1	2.874941	8
Qatar Grand Prix
HARD	1	3.271513	6
INTERMEDIATE	4	5.000000	6
MEDIUM	1	1.411722	5
SOFT	1	4.154930	6
Saudi Arabian Grand Prix
HARD	1	2.012779	4
INTERMEDIATE	1	1.000000	1
MEDIUM	1	1.072724	4
SOFT	1	3.304762	4
Singapore Grand Prix
HARD	1	1.988117	5
INTERMEDIATE	1	1.013149	4
MEDIUM	1	1.161593	4
SOFT	1	2.662222	5
Spanish Grand Prix
HARD	1	2.753523	6
INTERMEDIATE	5	5.500000	6
MEDIUM	1	2.248474	5
SOFT	1	1.638353	6
São Paulo Grand Prix
HARD	1	1.367347	5
INTERMEDIATE	1	1.693778	6
MEDIUM	1	2.456855	6
SOFT	1	3.526909	7
WET	2	2.000000	2
United States Grand Prix
HARD	1	2.325069	5
MEDIUM	1	1.188922	4
SOFT	1	2.228729	4

Feature Engineering

train <- train %>%
  mutate(
    TyreLife_x_Degradation = TyreLife * Cumulative_Degradation,

    Degradation_per_Lap = Cumulative_Degradation / (TyreLife + 1),
    
    TyreAge_Category = case_when(
      Compound == "SOFT"         & TyreLife > 15 ~ "old",
      Compound == "MEDIUM"       & TyreLife > 25 ~ "old",
      Compound == "HARD"         & TyreLife > 35 ~ "old",
      Compound == "INTERMEDIATE" & TyreLife > 20 ~ "old",
      TRUE ~ "fresh"
    )
  )

test <- test %>%
  mutate(
    TyreLife_x_Degradation = TyreLife * Cumulative_Degradation,

    Degradation_per_Lap = Cumulative_Degradation / (TyreLife + 1),
    
    TyreAge_Category = case_when(
      Compound == "SOFT" & TyreLife > 15 ~ "old",
      Compound == "MEDIUM" & TyreLife > 25 ~ "old",
      Compound == "HARD" & TyreLife > 35 ~ "old",
      Compound == "INTERMEDIATE" & TyreLife > 20 ~ "old",
      TRUE ~ "fresh"
    )
  )

train <- train %>%
  mutate(
    Losing_Positions = Position_Change < -1,
    
    InPoints = Position <= 10,

    Struggling_OldTyres = Losing_Positions & Degradation_per_Lap > median(Degradation_per_Lap, na.rm = TRUE)
  )

test <- test %>%
  mutate(
    Losing_Positions = Position_Change < -1,
    
    InPoints = Position <= 10,

    Struggling_OldTyres = Losing_Positions & Degradation_per_Lap > median(Degradation_per_Lap, na.rm = TRUE)
  )

Kaggle: S6 E5 Playground

2026-05-31

Predicting F1 Pit Stops

Missing Values

Data Exploration

Class Imbalance Check

Data Correlations

Tyre Life and Degradation

Pit Rate by Compound

Race Progress

Driver Position

Feature Engineering