I’ve always wondered and have never got a clear answer to this: is there an option to use ML with influxdb’s data AS the data is coming from say, IoT sensors, to predict / forecast data in real time, or is it only like we can use the data from influxdb after it comes in to train our model which is a separate process. Also would like to clarify if there’s any UI for influxdb v3 like influxdb v2 (local host 8086), because I can’t find to seem one but in one of your pages I’ve seen something similar to a UI.
Yes, you can write write Python program that does an ML job using your favote python framework or ml model with help of InfluxDB 3 Python Processing Engine. For example you can use trigger the WAL (as soon as data is written to InfluxDB 3) you can process the data and do anomaly detection using popular python library such as Prophet, see here for a blog post - Build a Time Series Forecasting Pipeline in InfluxDB 3 Without Writing Code | InfluxData about it.
We also have a new UI tool in beta that you can install separately and use it to do various operations on v3, see this page: Get started using InfluxDB 3 Explorer | InfluxDB 3 Explorer Documentation
Will definitely look into it. Thanks a lot!
1 Like
