如何用ggplot将多个函数整合到一张图表中？附数据集详情

Got it, let's break this down for you. You have a 25×6 dataframe with factor data (Mkt.RF, SMB, HML, RMW, CMA, WML), and want to combine multiple related plots or series into one visualization with ggplot2. Here's a step-by-step guide with concrete examples:

First: Prep Your Data (Critical!) #

ggplot2 works best with long-format data (one row per observation per variable) instead of your current wide format (one column per factor). Let's start by converting it using the tidyr package:

First, let's simulate a dataframe matching your structure (since your raw data was cut off):

# Load required packages
library(ggplot2)
library(tidyr)
library(dplyr)

# Simulate your 25x6 dataframe (replace this with your actual data)
set.seed(123)
df <- data.frame(
  Mkt.RF = rnorm(25),
  SMB = rnorm(25),
  HML = rnorm(25),
  RMW = rnorm(25),
  CMA = rnorm(25),
  WML = rnorm(25)
)
# Add a time/observation index (since you have 25 rows, assume this is time)
df$obs <- 1:nrow(df)

Now convert to long format:

df_long <- df %>%
  pivot_longer(cols = -obs, names_to = "Factor", values_to = "Return")

This gives you a dataframe with 3 columns: obs (your x-axis, e.g., time), Factor (the variable name), and Return (the value).

Option 1: Overlay All Series on One Plot #

If you want to compare all factors directly on the same axes, use color to distinguish them:

ggplot(df_long, aes(x = obs, y = Return, color = Factor)) +
  geom_line(linewidth = 1) +  # Use lines for time series
  geom_point(size = 2) +      # Add points for clarity
  labs(
    title = "Factor Returns Over Time",
    x = "Observation Number",
    y = "Return",
    color = "Factor"
  ) +
  theme_minimal() +
  scale_color_viridis_d(option = "plasma")  # Nice color palette for categorical variables

This will show all 6 factor series on one plot, each with a unique color.

Option 2: Faceted Plots (Separate Panels for Each Factor) #

If overlaying gets too cluttered, use facets to create a grid of small plots (one per factor):

ggplot(df_long, aes(x = obs, y = Return)) +
  geom_line(color = "#2c3e50", linewidth = 1) +
  geom_point(color = "#e74c3c", size = 2) +
  facet_wrap(~Factor, ncol = 2) +  # Arrange facets in 2 columns
  labs(
    title = "Factor Returns by Factor",
    x = "Observation Number",
    y = "Return"
  ) +
  theme_minimal() +
  theme(strip.background = element_rect(fill = "#f8f9fa"),
        strip.text = element_text(face = "bold"))

This keeps each factor's data separate but aligned for easy comparison.

Bonus: Adding Derived Functions (e.g., Rolling Averages) #

If you want to plot the original data alongside a derived function (like a rolling mean), calculate it first and merge it into your long data:

# Calculate rolling 3-period mean for each factor
df_rolling <- df_long %>%
  group_by(Factor) %>%
  mutate(Rolling_Mean = zoo::rollmean(Return, k = 3, fill = NA))  # Use zoo package for rolling stats

# Plot original returns + rolling mean
ggplot(df_rolling, aes(x = obs, y = Return)) +
  geom_line(color = "gray50", alpha = 0.6) +
  geom_line(aes(y = Rolling_Mean), color = "#2980b9", linewidth = 1.2) +
  facet_wrap(~Factor, ncol = 2) +
  labs(
    title = "Factor Returns with 3-Period Rolling Mean",
    x = "Observation Number",
    y = "Value"
  ) +
  theme_minimal()

This adds a smoothed rolling average line on top of each factor's raw return data.

Just replace the simulated data with your actual dataframe, and adjust the x-axis (if your rows are dates instead of observation numbers, use as.Date() to format it properly).

内容的提问来源于stack exchange，提问作者Mads