R-plotly子图仅显示底部图：复现mtcars数据缩放子图遇问题

Let's break down what's going wrong and how to fix it so your am==0 points show up in the top subplot as expected.

The Root Issue #

When you use as.integer(factor(am)), R's default factor ordering turns am=0 into 1 and am=1 into 2—that part is correct. The problem likely stems from either:

• Ambiguous subplot axis mapping in Plotly 4.9.0 (without explicit domain settings, subplots might render in an unexpected order)
• Using a continuous color scale for am (since it's numeric) instead of discrete colors, which makes it harder to verify group placement
• Missing explicit layout controls to anchor each subplot to its intended position

Solution 1: Explicit Subplot Creation (Most Intuitive) #

Split your data by am first, create separate plots for each group, then combine them. This eliminates confusion about axis mapping:

library(plotly)
library(dplyr)

# Split data into automatic (am=0) and manual (am=1) groups
mtcars_auto <- mtcars %>% filter(am == 0)
mtcars_manual <- mtcars %>% filter(am == 1)

# Create top subplot (am=0, automatic)
p_top <- plot_ly(mtcars_auto, 
                 x = ~mpg, 
                 y = ~qsec, 
                 color = ~factor(am, labels = c("Automatic", "Manual")),
                 type = "markers") %>%
  layout(yaxis = list(title = "qsec (Automatic)"))

# Create bottom subplot (am=1, manual)
p_bottom <- plot_ly(mtcars_manual, 
                    x = ~mpg, 
                    y = ~qsec, 
                    color = ~factor(am, labels = c("Automatic", "Manual")),
                    type = "markers") %>%
  layout(yaxis = list(title = "qsec (Manual)"))

# Combine subplots with shared X-axis, top-to-bottom order
subplot(p_top, p_bottom, nrows = 2, shareX = TRUE, titleY = TRUE)

Solution 2: Pipe-Friendly Fix with Explicit Axis Domains #

If you prefer keeping the single pipeline, fix the axis mapping and add explicit layout rules to enforce subplot positions:

library(plotly)
library(dplyr)

mtcars %>%
  # Directly map am to axis IDs (no factor ambiguity)
  mutate(axis_id = ifelse(am == 0, "y1", "y2")) %>%
  plot_ly(x = ~mpg, 
          y = ~qsec, 
          color = ~factor(am, labels = c("Automatic", "Manual")),
          yaxis = ~axis_id) %>%
  add_markers() %>%
  subplot(nrows = 2, shareX = TRUE) %>%
  # Define exact positions for each y-axis (top = y1, bottom = y2)
  layout(
    yaxis = list(domain = c(0.55, 1), title = "qsec (Automatic)"),
    yaxis2 = list(domain = c(0, 0.45), title = "qsec (Manual)")
  )

Key Fixes Explained #

• Discrete Color Mapping: Converting am to a factor (factor(am)) gives you distinct colors for each group, making it easy to confirm which points belong to which subplot.
• Explicit Axis Domains: The domain parameter in layout forces y1 (am=0) to occupy the top half of the plot and y2 (am=1) to occupy the bottom half.
• Clear Axis Mapping: Using ifelse instead of as.integer(factor(am)) removes any confusion about how am values map to axis IDs.

内容的提问来源于stack exchange，提问作者unamourdeswann