使用R Quantmod分析BCBA两只股票xts对象相关性遇报错求助

Hey there! Let's break down how to resolve this error and get your correlation analysis working smoothly.

Step 1: Fix the Cl() Usage (A Common Mistake) #

First off, that error Error in Cl("BCBA:GGAL") : subscript out of bounds probably comes from passing the ticker string directly to Cl() instead of the actual xts object. When you run getSymbols("BCBA:GGAL"), quantmod creates an xts object named BCBA.GGAL (colons get replaced with dots automatically). You need to use that object name with Cl(), not the original ticker string.

Step 2: Check and Rename Columns (If Needed) #

Sometimes data from sources like Google Finance comes with prefixed column names (e.g., BCBA.GGAL.Close instead of just Close). While Cl() usually detects columns containing "Close", if you still hit issues, you can explicitly rename columns to standard labels:

library(quantmod)

# Fetch the stock data
getSymbols("BCBA:FRAN", src = "google")
getSymbols("BCBA:GGAL", src = "google")

# Check current column names to confirm the structure
colnames(BCBA.FRAN)
colnames(BCBA.GGAL)

# Rename columns to standard quantmod-friendly names
colnames(BCBA.FRAN) <- c("Open", "High", "Low", "Close", "Volume", "Adjusted")
colnames(BCBA.GGAL) <- c("Open", "High", "Low", "Close", "Volume", "Adjusted")

Step 3: Extract Closing Prices and Merge Data #

Next, pull the closing prices from each xts object, then merge them to align dates (we’ll only keep dates where both stocks have valid data):

# Extract closing prices using the corrected xts objects
fran_close <- Cl(BCBA.FRAN)
ggal_close <- Cl(BCBA.GGAL)

# Merge the two time series (drop dates with missing data for either stock)
combined_prices <- merge(fran_close, ggal_close, all = FALSE)

# Rename merged columns for clarity
colnames(combined_prices) <- c("FRAN_Close", "GGAL_Close")

Step 4: Calculate Correlation #

Finally, compute the correlation between the two closing price series. We’ll use use = "complete.obs" to skip any missing values:

# Calculate Pearson correlation coefficient
correlation_value <- cor(combined_prices$FRAN_Close, combined_prices$GGAL_Close, use = "complete.obs")
cat("Correlation between FRAN and GGAL:", correlation_value, "\n")

# Or generate a full correlation matrix for the merged dataset
correlation_matrix <- cor(combined_prices, use = "complete.obs")
print(correlation_matrix)

Bonus: Visualize the Relationship #

If you want to see the relationship visually, add these plots:

# Plot both closing price series over time
plot(combined_prices, main = "FRAN vs GGAL Closing Prices (BCBA Exchange)")

# Scatter plot with a regression line to show the correlation trend
plot(combined_prices$FRAN_Close, combined_prices$GGAL_Close, 
     xlab = "FRAN Closing Price", ylab = "GGAL Closing Price", 
     main = "FRAN vs GGAL Correlation Scatter Plot")
abline(lm(combined_prices$GGAL_Close ~ combined_prices$FRAN_Close), col = "red", lwd = 2)

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