This is my working code for looking at velocity data from lab gravity currents obtained through PIV. At some point this might be stable but right now it is a work in progress!

Things to do:

• vorticity plots. basically port demo/plot.py into the plotter class

• can we identify qualitatively different regions of the flow?

  • do they have distinct pdfs?
  • does cantero have something to say here?

• DMD: can we recompose the flow from low order modes? are the stats the same?

More things to do:

• look at ogive plots

• distinguish between sampling dimension and time / space in front relative extraction.

• wavelet ensembles? can we increase confidence with more ensemble members

• make the pdf as a function of time and height work

  • plot with log height
  • plot for multiple ensembles

• distinguish ensembles - inter / intra run, inter parameter

• rapid distortion theory. eddy turnover time large compared with advective timescale?

• pdfs limited to particular events (eddies)

• fit log profile to vertical pdfs (log height)

• compute vertical pdfs with highlighted data exceeding certain percentile close together in space / time (i.e. same event)

Done:

• Non-dimensionalise

• interpolate zeros in pre processor - how does pandas do it?

• fit straight line / smooth front detection

  • check sensitivity of stats to different fittings

• overlay multiple runs to get first impression of similarity (use the single height over time)

• do the front relative transform along the other (space) axis and see what it looks like

• look at streamwise velocity

• subtract front speed to get front relative velocity

• is there a region of the flow in which the decomp mean front relative velocity is zero? (is this the region of statistical stationarity?)