Anomalize报错：未找到日期或日期时间列的问题咨询

Based on your data snippet, this error stems from two key issues:

• Your date column is stored as numeric Unix timestamps (not a proper Date or POSIXct datetime class that Anomalize recognizes)
• The dates are not sorted in chronological order, which is a requirement for time series anomaly detection tools like Anomalize.

Here's how to fix this step-by-step:

Step 1: Convert Numeric Timestamps to Proper Datetime Class #

Anomalize depends on columns explicitly marked as date/datetime types. Run this code to convert your numeric timestamps to a POSIXct object (your timestamps are in seconds since the Unix epoch):

# Convert numeric timestamp to POSIXct datetime
y$date <- as.POSIXct(y$date, origin = "1970-01-01")

# If you only need date (no time component), use as.Date instead:
# y$date <- as.Date(y$date, origin = "1970-01-01")

Step 2: Sort Your Data Chronologically #

Time series data must be ordered by date for Anomalize to process it correctly. Sort your data frame with:

y <- y[order(y$date), ]

Step 3: Verify the Fix #

Double-check that your date column is now the correct class and sorted:

# Check class of date column
class(y$date)
# Should return "POSIXct" "POSIXt" (or "Date" if you used as.Date)

# Confirm dates are sorted in ascending order
all(y$date == cummax(y$date))
# Should return TRUE if sorted correctly

Once you’ve completed these steps, re-run your Anomalize workflow—this should resolve the "no date or datetime column found" error.

内容的提问来源于stack exchange，提问作者Ted Mosby