olympicdash-py-2.qmd

---
title: "Olympic Games"
format: 
  dashboard:
    orientation: columns
    nav-buttons: [github]
    github: https://github.com/posit-conf-2024/olympicdash
logo: images/olympics-logo.svg
logo-alt: "Olympics logo with multicolored circles."
---

```{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 Summer season and non-NA medals
olympics = olympics_full[(olympics_full["season"] == "Summer") & (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["medal"] = pd.Categorical(olympics["medal"], categories = ["Bronze", "Silver", "Gold"])
```

## Column - Medals by sport and year

### Row - Medals by sport {height=60%}

```{python}
#| label: medals-by-sport
#| title: Medals by sport

# 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)) +
    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, 6.18)
    )
)
```

### Row - Medals by year {height=40%}

```{python}
#| label: medals-by-year
#| title: Medals by year

# 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)) +
    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, 3)
    )
)
```

## Column - Medals by country

```{python}
#| label: medals-by-country
#| title: Medals by country
# 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)
)
```