I am generating a heat map for spectral density data via this query:
SELECT “mag_vel” *1000 FROM “rft_data” WHERE (“mag_vel” < 0.01 AND “accId” = ‘X-1L39’) AND time >= now() - 30m GROUP BY “frq”
It’s 128 bins for a logarithmic (pre computed time series buckets) set of frequencies. It is all working great, but for larger time ranges it is getting very very slow. I am not sure how the data it preprocessed, but it seam all data is considered and some how put together. (good, but also very slow) as I have a spectra every 6s if logging is triggered. Up to 6 hours all works fine, but fetching a day or even more is nearly impossible even with patience. Is there any way I could pre process and mark (to only select) a sub set of “peak” data of do something a like?
See screenshot example, this is one of 6 panels total: