Goal of simplifying multi-color rasterize analysis into one line in hvplot?

[ahuang11]

I want to simplify this workflow down to a single call in hvplot.

df.hvplot(
    "LON",
    "LAT",
    groupby="BaseDateTime",
    geo=True,
    tiles=True,
    datashade=True,
    dynspread=True,
    by="VesselType",
    aggregator=ds.count_cat,
    cmap="glasbey_bw_minc_20_minl_30",
)

I think it's doable, but hvplot is missing a few features, so this issue is kind of like an epic

1. Automatic lon/lat to easting/northing Automatically or manually toggle to automatically use lon_lat_to_easting_northing #1375
2. Map categorical strings to color given a cmap (or default cmap) and use as color_key in datashade
3. Create legend based on the categories Util to create a custom legend holoviews#6320

For 2, I think rather than having a column be the actual color, color=["red", "green", "blue"], if it detects non-valid colors, it would automatically use a cmap to map categories to color (I always find that tedious) https://holoviews.org/user_guide/Style_Mapping.html#styling-mapping

To reproduce:
Data download: https://coast.noaa.gov/htdata/CMSP/AISDataHandler/2020/index.html

import pandas as pd
import hvplot.pandas
import geoviews as gv
import pandas as pd
import datashader as ds
import colorcet as cc
import holoviews as hv
import holoviews.operation.datashader as hd
from holoviews.util.transform import lon_lat_to_easting_northing

df = pd.read_csv("AIS_2020_01_01.csv", parse_dates=True, index_col="BaseDateTime")
df = df.assign(
    **pd.concat(lon_lat_to_easting_northing(df["LON"], df["LAT"]), axis=1).rename(
        {"LON": "EASTING", "LAT": "NORTHING"}, axis=1
    )
)

df.index = df.index.round("1Min")
df = df.sort_index()
df.index = df.index.astype(str)

vessel_types = pd.read_csv("AIS_categories.csv")

categories = {
    r.num: r.category if r.category in [0, 2, 3, 6, 7, 16, 14, 19, 12, 18] else 21
    for i, r in vessel_types.iterrows()
}


def category_desc(val):
    """Return description for the category with the indicated integer value"""
    return vessel_types[vessel_types.category == val].iloc[0].category_desc


groups = {categories[i]: category_desc(categories[i]) for i in categories.keys()}
colors = cc.glasbey_bw_minc_20_minl_30
color_key = {
    list(groups.keys())[i]: tuple(int(e * 255.0) for e in v)
    for i, v in enumerate(colors[: (len(groups))][::-1])
}
legend = hv.NdOverlay(
    {
        groups[k]: hv.Points([0, 0], label=str(groups[k])).opts(
            color=cc.rgb_to_hex(*v), size=0
        )
        for k, v in color_key.items()
    }
)

df = df.loc[df["VesselType"].isin(color_key)]

pts = hv.Dataset(df, kdims=["EASTING", "NORTHING"], vdims=["VesselType"]).to(
    hv.Points, groupby=["BaseDateTime"]
)
points = hd.dynspread(
    hd.datashade(pts, aggregator=ds.count_cat("VesselType"), color_key=color_key)
)

tiles = (
    hv.element.tiles.ESRI()
    .opts(alpha=0.4, bgcolor="black")
    .opts(responsive=True, min_height=600)
)
labels = hv.element.tiles.CartoDark().opts(alpha=0.7, level="glyph")
display(
    tiles
    * labels
    * points.opts(show_legend=False)
    * legend.opts(xaxis="bare", yaxis="bare", title="")
)