9  Trends

We will keep using the heart disease dataset (see ?sec-import_local). However, we will consider a small subset of these data in order to lower computing resources.

R

Code

df <- df |> group_by(heart_disease) |> slice_head(n = 50)

Python

Code

df = df.groupby("heart_disease").head(50)

9.1 Scatter Plots

9.1.1 A simple scatter plot

R

library(ggplot2)

df |> 
    ggplot(aes(bmi, sleep_time)) +
    geom_point()

Python

import matplotlib.pyplot as plt
import seaborn as sns
import seaborn.objects as so

sns.set_theme()

(
    so.Plot(df, "bmi", "sleep_time")
    .add(so.Scatter())
    .show()
)

9.1.2 Color by variable

R

df |> 
    ggplot(aes(bmi, sleep_time, color = heart_disease)) +
    geom_point(alpha = 0.6) +
    scale_color_brewer(palette = "Set1")

Python

(
    so.Plot(df, "bmi", "sleep_time", color="heart_disease")
    .add(so.Scatter(alpha=0.6))
    .scale(color="Set1")
    .show()
)

or also:

sns.relplot(
    x="bmi", y="sleep_time", hue="heart_disease",
    alpha=0.6,
    palette="Set1",
    data=df
);
plt.show()

9.2 Line Plots

R

df |> 
    ggplot(aes(bmi, sleep_time, color = heart_disease, linetype = smoking)) +
    geom_line(alpha = 0.6) +
    scale_color_viridis_d(option = "plasma")

Python

sns.lineplot(
    x="bmi", y="sleep_time", hue="heart_disease", style="smoking",
    alpha=0.6, 
    palette="plasma",
    data=df
)

9.3 Combine different layers

R

df |> 
    ggplot(aes(bmi, sleep_time, color = heart_disease)) +
    geom_line(aes(linetype = smoking), alpha = 0.6) +
    geom_point(aes(shape = smoking), alpha = 0.6) +
    scale_color_viridis_d(option = "plasma")

Python

sns.relplot(
    x="bmi", y="sleep_time", hue="heart_disease", style="smoking",
    kind="line", markers=True, alpha=0.6, 
    palette="plasma",
    data=df
);
plt.show()