如何在 R 中使用逻辑回归模型绘制线性图？

问：

我想确定我的连续变量和二元结果变量是否具有线性关系。我该怎么做？我尝试了几种不同的情节，但没有一个能很好地显示这种关系。我的数据如下所示：

         anxiety physical_activity
     <dbl>             <dbl>
 1       1               240
 2       0               210
 3       0                90
 4       0                60
 5       1               160
 6       0               300
 7       0                 0
 8       0                40
 9       0                30
10       1                60

我的模型如下所示：

model <- glm(anxiety ~ physical_activity_hours, data=data, family=binomial(), na.action = na.exclude)

我试过这个情节：

# Extracting residuals from the model
residuals <- resid(model)

# Creating the data frame for plotting
plot_data <- data.frame(physical_activity_hours = data$physical_activity_hours, residuals = residuals)

# Generating the plot
ggplot(plot_data, aes(x = physical_activity_hours, y = residuals)) +
  geom_point(alpha = 0.5) +  # Plot the raw data points with some transparency
  stat_smooth(span = 0.5, se = FALSE) +  # Add a loess smoothed line; adjust span as needed
  theme_minimal() +  # Use a minimal theme for a clean plot
  labs(x = "Physical Activity Hours", y = "Residuals", title = "Residuals vs. Physical Activity with Moving Average")

而这个情节：

> data %>%
+   mutate(anxiety = as.numeric(anxiety)) %>%
+   pivot_longer(cols = c("physical_activity_hours"), names_to = "predictors") %>%
+   ggplot(aes(x = value, y = anxiety)) + 
+   geom_point(size = 0.5, alpha = 0.5) +
+   geom_smooth(method = "loess") + 
+   facet_wrap(~predictors, scales = "free_x")

他们俩都不是很好。

我做错了什么？