Open NickMortimer opened 3 years ago
Here's Roland's implementation in numpy
I think we are looking at something like this with a user function
pings = xr.open_dataset('https://s3-ap-southeast-2.amazonaws.com/imos-data/IMOS/SOOP/SOOP-BA/VLMJ_Investigator/Investigator_20160218-20160221/IMOS_SOOP-BA_AE_20160218T232139Z_VLMJ_FV02_Investigator-EK60-18-38-70-120_END-20160221T084435Z_C-20160630T221214Z.nc',chunks={'TIME':100})
pings.Sv.rolling(DEPTH=10).construct("window_dim").reduce(np.nanmean,dim=('window_dim'))
It could also be good to run Roland's implementation on an example to verify our results
It could also be good to run Roland's implementation on an example to verify our results
Yes that would be great
https://besjournals.onlinelibrary.wiley.com/doi/full/10.1111/2041-210X.12396
Have a look at this and xarray
http://xarray.pydata.org/en/stable/generated/xarray.core.rolling.DataArrayRolling.construct.html