Discrete colour scale for lines in Plotly graph objects (Python)

Question

I'm mapping lines in plotly.graph_objects.Scattermapbox with fig.add_trace. Ideally, I would like to change the colour of the lines, such that it corresponds to an attribute in another column. For example:

Line	Attribute
LINE01	Highway
LINE02	River
LINE03	Highway

I can't figure out how to do this. When I create my plot, each line appears as a legend entry (LINE01 appears as trace 0, etc.) This is my code:

import plotly.graph_objects as go
fig = go.Figure()


for i in range(len(df)):
    fig.add_trace(go.Scattermapbox(
        mode = "lines",
        lon = [df['lon.start'][i], df['lon.end'][i]],
        lat = [df['lat.start'][i], df['lat.end'][i]],
        line = dict(width = 3)
        )
    )

How can I change it such that my legend is grouped by the Attribute column to create a discrete colour scale?

score 2 · Accepted Answer · answered Sep 30 '21 at 15:50

your question did not have sample data for roads and rivers. Have sourced UK rivers and roads from UK gov sources
there are two approaches to adding line layers to a plotly figure
1. as mapbox layers. This carries the advantage that you can utilise plotly and geopandas geojson capabilities to more simply use reference mapping information from 3rd party sources
2. as traces - the approach you have been using. You have used start and end of these lines which means you loose all coordinates between
I have used this answer How to plot visualize a Linestring over a map with Python? on how to generate lines on a mapbox figure.
have used a subset of data, just because figure generation time is significant with full set of rivers and roads

source sample rivers and roads data

import urllib
from pathlib import Path
from zipfile import ZipFile
import geopandas as gpd
import pandas as pd

# get some river and road geometry....
src = [
    {
        "name": "rivers",
        "suffix": ".shp",
        "color": "blue",
        "width": 1.5,
        "url": "https://environment.data.gov.uk/UserDownloads/interactive/023ce3a412b84aca949cad6dcf6c5338191808/EA_StatutoryMainRiverMap_SHP_Full.zip",
    },
    {
        "name": "roads",
        "suffix": ".shp",
        "color": "red",
        "width": 3,
        "url": "https://maps.dft.gov.uk/major-road-network-shapefile/Major_Road_Network_2018_Open_Roads.zip",
    },
]
data = {}
for s in src:
    f = Path.cwd().joinpath(urllib.parse.urlparse(s["url"]).path.split("/")[-1])
    if not f.exists():
        r = requests.get(s["url"],stream=True,)
        with open(f, "wb") as fd:
            for chunk in r.iter_content(chunk_size=128):
                fd.write(chunk)

    fz = ZipFile(f)
    fz.extractall(f.parent.joinpath(f.stem))

    data[s["name"]] = gpd.read_file(
        f.parent.joinpath(f.stem).joinpath(
            [
                f.filename
                for f in fz.infolist()
                if Path(f.filename).suffix == s["suffix"]
            ][0]
        )
    ).assign(source_name=s["name"])
gdf = pd.concat(data.values()).to_crs("EPSG:4326")

use mapbox layers

import plotly.graph_objects as go
import json

# let's work with longer rivers and smaller numer of random roads
gdf2 = gdf.loc[gdf["length_km"].gt(50).fillna(False) | gdf["roadNumber"].isin(gdf["roadNumber"].fillna("").sample(50).unique())  ]
fig = go.Figure(go.Scattermapbox())

# use geopandas and plotly geojson layer capabilities,  keep full definition of line strings
fig.update_layout(
    margin={"l": 0, "r": 0, "t": 0, "b": 0},
    mapbox={
        "style": "carto-positron",
        "zoom": 4,
        "center": {
            "lon": gdf.total_bounds[[0, 2]].mean(),
            "lat": gdf.total_bounds[[1, 3]].mean(),
        },
        "layers": [
            {
                "source": json.loads(gdf2.loc[gdf2["source_name"].eq(s["name"])].geometry.to_json()),
                "below": "traces",
                "type": "line",
                "color": s["color"],
                "line": {"width": s["width"]},
            }
            for s in src
        ],
    },
)

use mapbox lines

import numpy as np

# plotly takes array delimited with None between lines. Use numpy padding and shaping to generate this array
# from pair of features
def line_array(df, cols):
    return np.pad(
        df.loc[:, cols].values, [(0, 0), (0, 1)], constant_values=None
    ).reshape(1, (len(df) * 3))[0]


# map to question columns.... looses all detail of a linestring
gdf3 = gdf2.join(
    gdf2.geometry.bounds.rename(
        columns={
            "minx": "lon.start",
            "miny": "lat.start",
            "maxx": "lon.end",
            "maxy": "lat.end",
        }
    )
)

fig = go.Figure(
    [
        go.Scattermapbox(
            name=g[0],
            lat=line_array(g[1], ["lat.start", "lat.end"]),
            lon=line_array(g[1], ["lon.start", "lon.end"]),
            mode="lines",
        )
        for g in gdf3.groupby("source_name")
    ]
)
fig.update_layout(
    margin={"l": 0, "r": 0, "t": 15, "b": 0},
    mapbox={
        "style": "carto-positron",
        "zoom": 4,
        "center": {
            "lon": gdf3.loc[:, ["lon.start", "lon.end"]].mean().mean(),
            "lat": gdf3.loc[:, ["lat.start", "lat.end"]].mean().mean(),
        },
    },
)

Thank you for your answer. Do you mind sharing the ```head``` of ```gdf```, so I can see what it should look like? Thanks again — lizard, Oct 07 '21 at 10:09
you can get it by running the **sourcing** code included in the answer — Rob Raymond, Oct 07 '21 at 10:15

Discrete colour scale for lines in Plotly graph objects (Python)

1 Answers1

source sample rivers and roads data

use mapbox layers

use mapbox lines