i recently started using holoviz ecosystem a bit more, and i may have found something strange.
Maybe i’m not doing it right, feel free to correct me
(i did a notebook here explaining the problem if you want to see : holoviews_decimate_pb.ipynb
I want to plot some gv.Points of satellite data. The number of points to plot is usually huge, but i show a minimal example here.
I know there is some other way to plot huge data with datashader, but usually i like what decimated do.
import xarray as xr import geoviews as gv import holoviews as hv from holoviews.operation import decimate hv.extension('bokeh') print("holoviews : ", hv.__version__) print("geoviews : ", gv.__version__)
holoviews : 1.14.4
geoviews : 1.9.1
# free available data ds = xr.open_dataset("https://opendap.jpl.nasa.gov/opendap/hyrax/allData/jason3/preview/L2/GPS-OGDR/c100/JA3_GPSOPR_2PdS100_013_20181026_140651_20181026_160516.nc") da = ds.ssha.load() da
when i plot the data with gv.Points, i have expected results.
but when i use holoviews.operation.decimate, and want to plot np.datetime64 values it breaks down lon/lat format and datetime64
(i appologize for my paint skills)
Is this behavior intended ?