anomalize | Tidy anomaly detection | Predictive Analytics library

The time_decompose() function requires data in the form of:

A tibble or tbl_time object

(from ?time_decompose)

Perhaps zem is a data.frame? You can include as_tibble() in the pipe to make sure it is a tibble ahead of time.

In addition, it expects to work on time based data:

It is designed to work with time-based data, and as such must have a column that contains date or datetime information.

I added to your test data a column with dates. Here is a working example:

The anomolize documentation gives a great example of how to apply anomolize() to a time series

This generates the Remainder_L1 and Remainder_L2 values for CRAN tidyverse downloads (that data comes with the anomolize package, so no need to import data, just run the code below to see how it generates the columns

knitr::all_labels() returns NULL if it is not called in the full knitting process. When you run a single code chunk in the R Markdown Notebook, the document is not fully knitted.

To make knitr::all_labels() return all chunk labels, you should not use the notebook mode or preview the notebook, but knit the document instead (e.g., to the html_document output format).

There seems to be a bug when the name of the date column is start_date. Try renaming the column to date. I've opened up Issue 81 on GitHub.

Approach 2 should work as expected? The error message is related to the stl() that requires at least two seasonal periods to estimate. For example, daily data needs at least 14 observations for stl() to run. Increasing the window size .size = 7 * 3 works fine.

The problem is that your date variable is not unique.