PostgreSQL Query
Budget: $30 – $250 USD
I would like to represent the value shift on the timeseries as on the picture in the form of time-weighted average.
There are 3 columns in the table.
1. Vector -> X axis
2. Magic number -> Y axis
3. Time (time shift represents different colours)
I would like to create a solution preferably in the postgres/timescaledb that it would provide continues/timeseries value change of (both X and Y axis).
The correlation between X and Y should look at the 2 closest match vectors below 0 (Magic number) and 2 closest points above 0 from the wide range of vectors. (Total 4 reference vector points)
My assumptions to solve it (could be inaccurate/not sure)
Is to calculate the area below the curve of 4 Vector points and average that over time in 5 min buckets. Or create time-weighted average from with a time shift (timescaledb)
There are 3 columns in the table.
1. Vector -> X axis
2. Magic number -> Y axis
3. Time (time shift represents different colours)
I would like to create a solution preferably in the postgres/timescaledb that it would provide continues/timeseries value change of (both X and Y axis).
The correlation between X and Y should look at the 2 closest match vectors below 0 (Magic number) and 2 closest points above 0 from the wide range of vectors. (Total 4 reference vector points)
My assumptions to solve it (could be inaccurate/not sure)
Is to calculate the area below the curve of 4 Vector points and average that over time in 5 min buckets. Or create time-weighted average from with a time shift (timescaledb)