xbitinfo.graphics.add_bitinfo_labels#

xbitinfo.graphics.add_bitinfo_labels(da, info_per_bit, inflevels, ax=None, x_dim_name='lon', y_dim_name='lat', lon_coord_name='guess', lat_coord_name='guess', label_latitude='center', label_latitude_offset=8, **kwargs)[source]#

Helper function for visualization of Figure 3 in Klöwer et al. 2021. Adds latitudinal lines and labels with keepbits and information content for each slice.

Klöwer, M., Razinger, M., Dominguez, J. J., Düben, P. D., & Palmer, T. N. (2021). Compressing atmospheric data into its real information content. Nature Computational Science, 1(11), 713–724. doi: 10/gnm4jj

Parameters

  • da (xarray.DataArray()) – Plotted data

  • info_per_bit (dict) – Information content of each bit for each variable in da. This is the output from xbitinfo.xbitinfo.get_bitinformation().

  • inflevels (list of floats) – Level of information that shall be preserved.

  • ax (plt.Axes or None) – Axes. If None, get current axis.

  • x_dim_name (str) – Name of the x dimension. Defaults to "lon".

  • y_dim_name (str) – Name of the y dimension. Defaults to "lat".

  • lon_coord_name (str) – Name of the longitude coordinate. Only matters when plotting with multi-dimensional coordinates (i.e. curvilinear grids) with cartopy (when transform=ccrs.Geodetic() must be also set via kwargs). Defaults to x_dim_name.

  • lat_coord_name (str) – Name of the latitude coordinate. Only matters when plotting with multi-dimensional coordinates (i.e. curvilinear grids) with cartopy (when transform=ccrs.Geodetic() must be also set via kwargs). Defaults to y_dim_name.

  • label_latitude (float or str) – Latitude for the label. Defaults to "center", which uses the mean lat_coord_name.

  • label_latitude_offset (float) – Distance between keepbits = int and x% label. Defaults to 8.

  • kwargs (dict) – Kwargs to be passed to ax.text and ax.plot. Use transform=ccrs.Geodetic() when using cartopy

Example

Plotting a single-dimension coordinate dataset: >>> ds = xr.tutorial.load_dataset(“air_temperature”) >>> info_per_bit = xb.get_bitinformation(ds, dim=”lon”) >>> inflevels = [1.0, 0.9999, 0.99, 0.975, 0.95] >>> ds_bitrounded_along_lon = xb.bitround.bitround_along_dim( … ds, info_per_bit, dim=”lon”, inflevels=inflevels … ) >>> diff = (ds - ds_bitrounded_along_lon)[“air”].isel(time=0) >>> diff.plot() # doctest: +ELLIPSIS <matplotlib.collections.QuadMesh object at …> >>> add_bitinfo_labels(diff, info_per_bit, inflevels) # doctest: +ELLIPSIS

Plotting a multi-dimensional coordinate dataset >>> v = “Tair” >>> ds = xr.tutorial.load_dataset(“rasm”) >>> dim = “y” >>> info_per_bit = xb.get_bitinformation(ds, dim=dim) >>> ds_bitrounded_along_lon = xb.bitround.bitround_along_dim( … ds, info_per_bit, dim=dim, inflevels=inflevels … ) >>> import cartopy.crs as ccrs # doctest: +SKIP >>> fig, axis = plt.subplots( # doctest: +SKIP … 1, 1, subplot_kw=dict(projection=ccrs.PlateCarree()) … ) >>> (ds - ds_bitrounded_along_lon)[v].isel(time=-10).plot( … ax=axis, transform=ccrs.PlateCarree() … ) # doctest: +SKIP >>> add_bitinfo_labels( … (ds - ds_bitrounded_along_lon)[v].isel(time=0), … lon_coord_name=”xc”, … lat_coord_name=”yc”, … x_dim_name=”x”, … y_dim_name=”y”, … transform=ccrs.Geodetic(), … ) # doctest: +SKIP