如何用R读取OpenDAP输出的四维网格ASCII文件并按时间和深度拆分生成XYZ数据框

Got it, let's walk through how to get your OpenDAP ASCII grid data into R and split it into the XYZ data frames you need—one for each time and depth combination. Here's a step-by-step approach:

First, we'll read the entire file into R as a character vector. This lets us handle the custom dimension formatting at the start of each line easily.

# Replace "your_data_file.txt" with the actual path to your file
raw_data <- readLines("your_data_file.txt")

Each line starts with a dimension tag like [0][0][0] (representing time, depth, latitude) followed by comma-separated values for each longitude. We'll use the stringr package to extract these components and turn each line into a small data frame, then combine everything into one big table.

# Install stringr if you haven't already
if (!require(stringr)) {
  install.packages("stringr")
  library(stringr)
}

# Function to parse a single line into a structured row set
parse_line <- function(line) {
  # Pull out the dimension tag (e.g., "[0][0][0]")
  dim_tag <- str_extract(line, "^\\[.*?\\]")
  # Extract the numeric values for time, depth, latitude
  dim_values <- as.integer(str_extract_all(dim_tag, "\\d+")[[1]])
  names(dim_values) <- c("time", "depth", "latitude")
  
  # Extract the longitude-wise data values
  value_list <- as.integer(str_split(str_remove(line, "^\\[.*?\\], "), ", ")[[1]])
  # Create a longitude sequence (assuming it starts at 1 and increments by 1; adjust if needed)
  longitude_seq <- seq_along(value_list)
  
  # Combine into a data frame
  data.frame(
    time = dim_values["time"],
    depth = dim_values["depth"],
    latitude = dim_values["latitude"],
    longitude = longitude_seq,
    sst = value_list # Using "sst" as in your example
  )
}

# Apply the function to all lines and combine into one data frame
full_structured_df <- do.call(rbind, lapply(raw_data, parse_line))

Note: If your longitude values aren't just sequential integers (1,2,3...), replace longitude_seq <- seq_along(value_list) with your actual longitude vector (e.g., a pre-defined list of longitude coordinates matching the number of values per line).

Now we'll split the full data frame into individual XYZ tables, grouped by each unique time + depth combination. We'll also name the resulting list elements for easy access.

# Split the data frame by time and depth
xyz_dfs <- split(full_structured_df[, c("latitude", "longitude", "sst")],
                 list(full_structured_df$time, full_structured_df$depth))

# Rename list elements to be more readable (e.g., "time_0_depth_0")
names(xyz_dfs) <- paste0("time_", 
                         sapply(strsplit(names(xyz_dfs), "\\."), `[`, 1),
                         "_depth_",
                         sapply(strsplit(names(xyz_dfs), "\\."), `[`, 2))

# Example: Check the first time + depth combination
head(xyz_dfs$time_0_depth_0)

This will give you a list where each element is an XYZ data frame exactly like your example—with columns for latitude, longitude, and sst—one for every time and depth pair in your original data.

内容的提问来源于stack exchange，提问作者Eko Susilo