Kats, a kit to analyze time series data, a lightweight, easy-to-use, generalizable, and extendable framework to perform time series analysis, from understanding the key statistics and characteristics, detecting change points and anomalies, to forecasting future trends.
MIT License
4.88k
stars
534
forks
source link
Non-equally distant point in time for anomaly detection #245
I was trying to use Kats MultivariateAnomalyDetector in a multivariate dataset that has points in time not equally distributed, i.e. the distance in time is variable.
I notice that an error is raised, since it requires that distance is the same for all points. Is there a solution that not involve the modification of dataset to make its points equally distributed in time but rather adapt the algorithm to handle points that has different distance in time? I read the documentation but unfortunately I found nothing.
From what I know Facebook Prophet can do this, but cannot handle multivariate dataset (correct me if I'm wrong).
I was trying to use Kats MultivariateAnomalyDetector in a multivariate dataset that has points in time not equally distributed, i.e. the distance in time is variable.
I notice that an error is raised, since it requires that distance is the same for all points. Is there a solution that not involve the modification of dataset to make its points equally distributed in time but rather adapt the algorithm to handle points that has different distance in time? I read the documentation but unfortunately I found nothing.
From what I know Facebook Prophet can do this, but cannot handle multivariate dataset (correct me if I'm wrong).
Thanks in advance