Fourier Transform

(Nithin Devang) #1

Previously I was using mysql and php/google charts to store data and I shifted to elasticsearch. I was using custom php function for fourier transform and display spectral data (chart). Is there anyway I can build a spectral analysis (Fast Fourier Transform) chart using Kibana for data present in Elastic Search?

(Brandon Kobel) #2

@Nithin_Devang all the visualizations in Kibana operate off of Elasticsearch aggregates, and from my understanding Fourier Transforms the discrete underlying data point so Kibana's visualizations likely aren't going to help you.

It'd be theoretically possible for you to store and retrieve the raw data in Elasticsearch, apply the Fourier Transform and then 're-index' your results in Elasticsearch for display in Kibana.

(Nithin Devang) #3

To create a Fourier transform I need set of historical data and I may not be able to convert it in real time.

(Brandon Kobel) #4

To create a Fourier transform I need set of historical data and I may not be able to convert it in real time.

Sorry, I'm not following.

(Nithin Devang) #5

Ok. The thing is I am getting real time data from sensors and adding that to Elasticsearch index. This is a time series data and you can plot across x axis as time series and values and y axis and this graph look good and one value is not dependent on previous value for any calculation.
But when it comes to Fast Fourier transform I cannot read a live sensor data and do a Fourier transform on that and add to Elasticsearch. Fourier transform formula takes a set of all historical data to calculate Frequency. So I am not sure how can I store and retrieve by applying Fourier transform on individual values.

(Brandon Kobel) #6

I think I'm following now, the point that I was trying to make earlier is that Kibana's current visualizations utilize aggregations, so you won't be able to get all historical data to perform the fourier transform with any of the existing visualizations. It's possible that you could write your own plugin/visualization for Kibana that does so, but we don't have any built-in capabilities that will make this possible.

(system) #7

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.