使用R语言quantmod库读取CSV时如何保留字符型行名

Got it, let's tackle this issue— I’ve dealt with this exact problem when working with quantmod and date-formatted row names in CSVs. Here’s why it’s happening and how to fix it:

Why Your Row Names Are Turning to 1, 2, 3... #

The most common culprit is that quantmod’s default CSV-reading functions (like read.zoo or getSymbols with src="csv") aren’t recognizing your date strings as the index/row names of your dataset. Instead, they’re auto-generating integer row IDs (starting at 1) and treating your dates as just another data column (or ignoring them entirely).

Step-by-Step Fixes #

Option 1: Read with read.csv First, Then Convert to xts #

This is the most straightforward approach for preserving row names:

1. Use base R’s read.csv to load the data, explicitly telling it to use the first column as row names:

   # Load the CSV, keep first column as row names
   raw_data <- read.csv("your_data.csv", row.names = 1, stringsAsFactors = FALSE)
   

2. Convert the data frame to an xts object (quantmod’s preferred format) while parsing the row names as dates:

   library(quantmod)
   # Convert row names to Date objects and set as the xts index
   quantmod_data <- as.xts(raw_data, order.by = as.Date(rownames(raw_data)))
   

Option 2: Use read.zoo Directly with Proper Arguments #

Since quantmod relies on zoo/xts, you can use read.zoo directly and specify how to handle your date row names:

library(quantmod)
# Read CSV, use first column as row names, parse them as dates
zoo_data <- read.zoo(
  "your_data.csv",
  header = TRUE,          # If your CSV has column headers
  sep = ",",              # Match your CSV's delimiter
  row.names = 1,          # Use first column as row names
  FUN = as.Date           # Convert row names to Date objects
)
# Convert to xts for quantmod compatibility
quantmod_data <- as.xts(zoo_data)

Option 3: Fix getSymbols for CSV Reads #

If you’re using getSymbols to load your CSV, specify the index column explicitly:

library(quantmod)
# Tell getSymbols to use the first column as the date index
getSymbols(
  "your_data.csv",
  src = "csv",
  index.column = 1,       # Set first column as the time index
  header = TRUE,
  sep = ","
)

Key Notes #

• Double-check your CSV structure: Make sure your date strings are in the first column of the file. If the column has a header (like "Date"), use index.column = "Date" instead of 1 in the above functions.
• For non-standard date formats (e.g., 08/05/1970 instead of 1970-05-08), adjust the FUN argument to parse correctly:

  # Example for DD/MM/YYYY format
  FUN = function(x) as.Date(x, format = "%d/%m/%Y")
  

Give these methods a shot, and your original date row names should stay intact instead of being replaced with numeric IDs!

内容的提问来源于stack exchange，提问作者HonnSolo