Exploratory Community

Bug in anomaly detection

#1

Hi, I’m doing a time series research and i think I found a bug.
After setting up my series, I can do anomaly detection with seem to work fine.
The problem arises when selecting “Direction of anomaly” to Positive. (by default is positive and negative)
Now click apply and run analysis againg and you get

Error in .f(.x[[i]], …) : object ‘negative_anomalies’ not found

I am assuming Direction of anomaly just sets the analysis to detect only positive values and ignore negatives.

Thanks and good job for this great piece of software!

#2

Hi Javier,

Thank you for reporting this issue.
It indeed seems like a bug.
I will work on fixing it

Thank you,
Hideaki