如何用循环统计数据框wb各列NA值并填充至向量naVals

Hey there! Let's troubleshoot why your loop isn't working to tally up NA values per column in your wb dataframe and store those counts in the naVals vector. Even though there's a simpler vectorized approach (colSums(is.na(wb))), I get wanting to nail the loop for learning purposes—let's break this down.

Common Mistakes That Break Your Loop #

First, let's cover the most frequent missteps that cause this kind of loop to fail:

• Not initializing the naVals vector properly: If you start with an empty vector (naVals <- c()), trying to assign to naVals[i] will throw an "subscript out of bounds" error because the vector doesn't have space for that index yet.
• Using row indices instead of column indices: Accidentally looping over 1:nrow(wb) instead of 1:ncol(wb) will count row-wise NAs instead of column-wise (or just break entirely if row/column counts differ).
• Incorrectly selecting columns: Using wb[i] instead of wb[,i]—for data frames, wb[i] returns a sub-dataframe with the i-th column, while wb[,i] returns the column as a vector (which is what you need for is.na() to work correctly).

The Working Loop Implementation #

Here's a corrected version that should run smoothly:

# Initialize the vector with a fixed length matching the number of columns in wb
naVals <- numeric(ncol(wb))

# Loop through each column index
for (i in 1:ncol(wb)) {
  # Calculate NA count for the i-th column and assign to naVals
  naVals[i] <- sum(is.na(wb[, i]))
}

Let's Walk Through What This Does #

1. Initialization: numeric(ncol(wb)) creates an empty numeric vector with exactly as many slots as there are columns in wb. This prevents the "subscript out of bounds" error you might have hit before.
2. Loop Range: 1:ncol(wb) ensures we iterate over every column in the dataframe, not rows.
3. NA Count Calculation:
   • wb[, i] selects the entire i-th column as a vector.
   • is.na(wb[, i]) converts every value in the column to TRUE if it's NA, FALSE otherwise.
   • sum(...) counts the number of TRUE values (since R treats TRUE as 1 and FALSE as 0 under the hood).

Example of What Might Have Gone Wrong #

If your original code looked something like this, here's why it failed:

# ❌ Wrong: Empty initial vector causes index errors
naVals <- c()
for (i in 1:ncol(wb)) {
  naVals[i] <- sum(is.na(wb[, i]))
}

# ❌ Wrong: Looping over rows instead of columns
naVals <- numeric(nrow(wb))
for (i in 1:nrow(wb)) {
  naVals[i] <- sum(is.na(wb[i, ]))
}

# ❌ Wrong: Incorrect column selection
naVals <- numeric(ncol(wb))
for (i in 1:ncol(wb)) {
  naVals[i] <- sum(is.na(wb[i]))
}

Once you fix those issues, your loop should work exactly as intended to populate naVals with column-wise NA counts.

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