olympicdash-1.qmd

---
title: "Olympic Games"
format: html
---

```{python}
#| label: load-packages
from plotnine import *
import great_tables as gt
import pandas as pd
```

```{python}
#| label: load-data
olympics_full = pd.read_csv("data/olympics.csv", low_memory=False)
```

```{python}
#| label: prep-data
# Filter for non-NA medals
olympics = olympics_full.loc[olympics_full["medal"].notna()]

# Split the team column at "-" into two columns
split_data = olympics["team"].str.split("-", n=1, expand=True)
olympics.loc[:, "team"] = split_data[0]

# Reorder the medal column categories
olympics.loc[:, "medal"] = pd.Categorical(
    olympics["medal"], categories=["Bronze", "Silver", "Gold"]
)
```

```{python}
#| label: summmer-prep-data
olympics = olympics[(olympics["season"] == "Summer")]
olympics.reset_index(drop=True, inplace=True)
```

## Medals by sport

```{python}
#| warning: false
# Lump the sport column to top 15 categories, grouping others as Other
top_15_sports = olympics["sport"].value_counts().nlargest(15).index
olympics["sport"] = olympics["sport"].apply(
    lambda x: x if x in top_15_sports else "Other"
)

# Convert the sport column to a categorical type with order based on frequency, and reverse the order
olympics["sport"] = pd.Categorical(
    olympics["sport"], categories=olympics["sport"].value_counts().index[::-1]
)

# Move the Other category of the sport column to the beginning
new_order = ["Other"] + [
    cat for cat in olympics["sport"].cat.categories if cat != "Other"
]
olympics["sport"] = olympics["sport"].cat.reorder_categories(new_order)

# Plot
(
    ggplot(olympics, aes(x="sport", fill="medal"))
    + geom_bar()
    + coord_flip()
    + guides(fill=guide_legend(reverse=True))
    + scale_fill_manual(
        values={"Gold": "#d4af37", "Silver": "#c0c0c0", "Bronze": "#cd7f32"}
    )
    + labs(x="", y="", fill="Medal")
    + theme_minimal()
    + theme(
        legend_position="inside",
        legend_position_inside=(0.9, 0.2),
        legend_direction="horizontal",
        legend_background=element_rect(fill="white", color="gray"),
        figure_size=(10, 5),
    )
)
```

## Medals by year

```{python}
# Count the occurrences of each medal per year
olympics_count = (
    olympics.groupby(["year", "medal"], observed=True).size().reset_index(name="n")
)

# Plot
(
    ggplot(olympics_count, aes(x="year", y="n", color="medal"))
    + geom_point(size=0.5)
    + geom_line()
    + guides(color=guide_legend(reverse=True))
    + scale_x_continuous(breaks=range(1896, 2020, 8))
    + scale_color_manual(
        values={"Gold": "#d4af37", "Silver": "#c0c0c0", "Bronze": "#cd7f32"}
    )
    + scale_fill_manual(
        values={"Gold": "#d4af37", "Silver": "#c0c0c0", "Bronze": "#cd7f32"}
    )
    + labs(x="Year", y="", color="Medal")
    + theme_minimal()
    + theme(
        legend_position="inside",
        legend_position_inside=(0.9, 0.2),
        legend_direction="horizontal",
        legend_background=element_rect(fill="white", color="gray"),
        figure_size=(10, 2.5),
    )
)
```

## Medals by country

```{python}
# Count the occurrences of each medal per team
olympics_count = olympics.groupby(["team", "medal"]).size().reset_index(name="n")

# Pivot olympics_count to get medals as columns
olympics_pivot = olympics_count.pivot_table(
    index="team", columns="medal", values="n", fill_value=0
)

# Calculate the total number of medals
olympics_pivot["Total"] = olympics_pivot[["Bronze", "Gold", "Silver"]].sum(axis=1)

# Reset the index and rearrange columns
olympics_pivot = olympics_pivot.reset_index()
olympics_pivot = olympics_pivot[["team", "Gold", "Silver", "Bronze", "Total"]]

# Sort by Total medals, then team
olympics_pivot = olympics_pivot.sort_values(
    by=["Total", "team"], ascending=[False, True]
)

# Rename the team column to Team
olympics_pivot.rename(columns={"team": "Team"}, inplace=True)

(gt.GT(olympics_pivot))
```