I need to compress data using a wavelet transform for a specific window. store coefficients and reconstruct data. how to proceed
Thanks for your question. I would probably use the Python CL + PyWavelets to read, transform, and then write data back into InfluxDB. You could also use the exec telegraf plugin to execute that python script if you need to run in periodically.
I needed a little more info:
Say I take a window of 300 samples(300 seconds of data), I then apply wavelet transform to that and I get my wavelet co - efficients ( say 200 values) associated with that window, now I have transformed timeseries to some other format but my data is for that window.
how would I apply a complex reconstruction algorithm that would fetch coefficients for individual windows,
recreate time series data from it and then display it on chronograf. How would the data storage architecture be for this case.
I don’t know enough about what you’re trying to accomplish or the parameters of your problem to give you a good answer…but I can make a guess. Assuming you are performing this analysis on metrics (and not events) and you’re working with regular time series, you could store those coefficients back into Influxdb for each window. Then you could perform your data transformation after fetching those parameters from InfluxDB for the particular window. The exec telegraf plugin executes the commands on every interval and parses metrics from their output in any one of the accepted Input Data Formats. Does that help?