咨询：基于Pima数据集实现Triceps与多因变量的简单线性回归绘图

It looks like you're aiming to create three separate scatter plots, each showing the linear relationship between triceps (on the x-axis) and one of pregnant, glucose, or pressure (on the y-axis) using the Pima dataset. Here's a complete, cleaned-up version of your code that finishes the first plot and adds the other two, plus some practical best practices:

# Load required libraries
library(ggplot2)
library(pdp) # Retained as per your original code, even if not used directly here

# Prepare dataset by removing rows with missing values
pima_clean <- na.omit(pima)
head(pima_clean)

# Plot 1: Triceps vs Number of Pregnancies
lin1 <- lm(triceps ~ pregnant, data = pima_clean)
coef1 <- coef(lin1)

p1 <- ggplot(data = pima_clean, aes(x = pregnant, y = triceps)) +
  geom_point(alpha = 0.5) + # Add alpha to reduce overplotting
  geom_abline(color = "blue", intercept = coef1[1], slope = coef1[2], size = 1) +
  labs(
    title = "Triceps Skin Fold vs Number of Pregnancies",
    x = "Number of Pregnancies",
    y = "Triceps Skin Fold Thickness (mm)"
  ) +
  theme_minimal()

# Plot 2: Triceps vs Glucose Level
lin2 <- lm(triceps ~ glucose, data = pima_clean)
coef2 <- coef(lin2)

p2 <- ggplot(data = pima_clean, aes(x = glucose, y = triceps)) +
  geom_point(alpha = 0.5) +
  geom_abline(color = "red", intercept = coef2[1], slope = coef2[2], size = 1) +
  labs(
    title = "Triceps Skin Fold vs Plasma Glucose Concentration",
    x = "Glucose Level (mg/dL)",
    y = "Triceps Skin Fold Thickness (mm)"
  ) +
  theme_minimal()

# Plot 3: Triceps vs Blood Pressure
lin3 <- lm(triceps ~ pressure, data = pima_clean)
coef3 <- coef(lin3)

p3 <- ggplot(data = pima_clean, aes(x = pressure, y = triceps)) +
  geom_point(alpha = 0.5) +
  geom_abline(color = "green", intercept = coef3[1], slope = coef3[2], size = 1) +
  labs(
    title = "Triceps Skin Fold vs Diastolic Blood Pressure",
    x = "Blood Pressure (mm Hg)",
    y = "Triceps Skin Fold Thickness (mm)"
  ) +
  theme_minimal()

# Optional: Arrange all plots in a grid for easy comparison
# install.packages("gridExtra") if you don't have it
library(gridExtra)
grid.arrange(p1, p2, p3, ncol = 2)

Quick Tips: #

• I renamed the cleaned dataset to pima_clean to avoid overwriting the raw original data—this is a safe practice for reproducible analysis.
• Swapped pima$column for unquoted column names inside aes()—this is the standard ggplot2 syntax and keeps code cleaner.
• Added alpha = 0.5 to points to make overcrowded areas easier to interpret, showing data density instead of overlapping dots.
• Included descriptive labels and titles to make your plots self-explanatory for anyone viewing them.

内容的提问来源于stack exchange，提问作者Ville