R语言初学者求助：na.omit()删除NA值全量删行、自定义筛选无效的问题排查

Hey there! Let's figure out why your NA removal attempts aren't working as expected. I'll break down both problems and walk you through fixes:

1. Why na.omit() is deleting all rows #

When na.omit() wipes out every single observation, it usually boils down to one of two scenarios:

• Every row has at least one standard NA: If even one column has NAs in every row, or every individual row has an NA in some column, na.omit() will drop all rows—since it removes any row containing any NA value.
• Your "missing values" aren't standard R NAs: Sometimes data uses placeholders like blank strings (""), the character "NA", or other non-standard values that is.na() doesn't recognize as true NAs. na.omit() only targets R's native NA type, so it won't touch these.

How to diagnose and fix this: #

First, check where the NAs (if any) are in your dataset:

# Total number of standard NAs across all columns
sum(is.na(Year2021_Trips_v2))

# Count NAs per column to see which ones are problematic
colSums(is.na(Year2021_Trips_v2))

If you see zero NAs here but know there are missing values, convert those non-standard placeholders to true NAs first:

# Turn blank strings into NAs
Year2021_Trips_v2[Year2021_Trips_v2 == ""] <- NA

# Turn character "NA" strings into NAs
Year2021_Trips_v2[Year2021_Trips_v2 == "NA"] <- NA

Then you can either re-run na.omit(), or use a more targeted approach to only remove NAs from specific columns (instead of dropping rows with any NA):

# Using dplyr to keep rows without NAs in key columns
Year2021_Trips_v2 <- Year2021_Trips_v2 %>%
  filter(!is.na(name_start_station), !is.na(ride_length)) # Add other columns as needed

2. Why your second code didn't remove any empty values #

The code you wrote:

Year2021_Trips_v2 <- Year2021_Trips[!(Year2021_Trips$name_start_station == "HQ QR" | Year2021_Trips$ride_length < 0),]

doesn't include any logic to handle NA values at all! It only filters out rows where name_start_station is "HQ QR" or ride_length is negative. To add NA removal to this filter, you need to explicitly check for NAs:

# Base R version combining your rules with NA checks
Year2021_Trips_v2 <- Year2021_Trips[!(Year2021_Trips$name_start_station == "HQ QR" | 
                                       Year2021_Trips$ride_length < 0 |
                                       is.na(Year2021_Trips$name_start_station) |
                                       is.na(Year2021_Trips$ride_length)),]

# Cleaner dplyr version
Year2021_Trips_v2 <- Year2021_Trips %>%
  filter(name_start_station != "HQ QR",
         ride_length >= 0,
         !is.na(name_start_station),
         !is.na(ride_length))

This way, you're combining your original filtering rules with checks to exclude rows with NAs in the columns you care about.

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