ggplot2 aes(fill)填充异常求助：足球赛程可视化颜色不符预期

Let's work through why your tile fill colors aren't behaving as expected and fix it step by step:

First, Diagnose the Core Issue #

Looking at your code, you're using fill=Pos where Pos is an integer representing league position. By default, ggplot treats integer values as continuous variables, so it applies a smooth gradient color scale. This likely doesn't match what you want—you probably intended each distinct league position to have a unique, discrete color instead of a blended gradient.

Also, quick check of your sample data: most teams have consistent Pos values across game weeks (GW), but if you meant for every tile of a team to use the same fixed position (like their current season-long rank), we can adjust for that too.

Solution 1: Treat Position as a Discrete Category #

Convert Pos to a factor so ggplot recognizes it as distinct ranks, then use a discrete color scale for clear differentiation:

library(tidyverse)

df %>% 
  ggplot() + 
  # Convert Pos to factor to get unique colors per rank
  geom_tile(aes(x = GW, y = team, fill = factor(Pos)), colour = "black") + 
  geom_text(aes(x = GW, y = team, label = oppo), size = 3) + 
  theme_void() + 
  theme(
    axis.text = element_text(face = "bold"),
    axis.text.y = element_text(margin = margin(0, -20, 0, 0))
  ) + 
  scale_x_continuous(position = "top", breaks = 1:15) + 
  labs(
    caption = paste("xxx rocks | ", Sys.Date(), sep = ""),
    fill = "League Position" # Add a clear legend title
  ) + 
  # Use a colorblind-friendly discrete palette; reverse so top ranks stand out more
  scale_fill_viridis_d(option = "plasma", reverse = TRUE)

Why this works: #

• factor(Pos) turns your integer ranks into distinct categories, so each position gets its own unique color instead of a gradient.
• scale_fill_viridis_d() uses a colorblind-friendly palette that makes ranks easy to tell apart, and reverse=TRUE ensures higher positions (smaller numbers) have more prominent colors—intuitive for league rankings.

Solution 2: Fix Positions to a Single Value Per Team (If Needed) #

If you want every tile for a team to use the same position color (e.g., their latest league rank), first standardize the position data:

# Get the latest position for each team (using the highest GW in your dataset)
team_fixed_pos <- df %>% 
  group_by(team) %>% 
  filter(GW == max(GW)) %>% 
  select(team, fixed_pos = Pos)

# Merge this fixed position back to your original data
df_updated <- df %>% 
  left_join(team_fixed_pos, by = "team")

# Plot using the unified fixed position
df_updated %>% 
  ggplot() + 
  geom_tile(aes(x = GW, y = team, fill = factor(fixed_pos)), colour = "black") + 
  geom_text(aes(x = GW, y = team, label = oppo), size = 3) + 
  theme_void() + 
  theme(
    axis.text = element_text(face = "bold"),
    axis.text.y = element_text(margin = margin(0, -20, 0, 0))
  ) + 
  scale_x_continuous(position = "top", breaks = 1:15) + 
  labs(
    caption = paste("xxx rocks | ", Sys.Date(), sep = ""),
    fill = "Latest League Position"
  ) + 
  scale_fill_viridis_d(option = "plasma", reverse = TRUE)

Why this works: #

• We first create a lookup table of each team's most recent position, then merge it with the original data. Now every tile for a team uses the same fixed position color, consistent across all game weeks.

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