Filling country colours by value of column in other DataFrame

Hi,

I’ve just started experimenting with Altair, and am loving it so far. I am, however, struggling somewhat with generating map plots.

Problem

I’m trying to plot a world map of confirmed Covid-19 cases, using data provided by John Hopkins. I want to colour countries by the number of confirmed cases, but have been unsuccessful.

Data

The code to retrieve and clean the data is:

import janitor # pip install pyjanitor
import pandas as pd

CASES_WORLDWIDE = "https://raw.githubusercontent.com/CSSEGISandData/COVID-19/web-data/data/cases_country.csv"

def get_worldwide_cases(url: str = CASES_WORLDWIDE):
    cases = pd.read_csv(url)
    cleaned = (
        cases.clean_names()
        .rename_column("long_", "lon")
        .transform_column("last_update", lambda x: pd.to_datetime(x).normalize())
    )
    return cleaned

world_source = get_worldwide_cases()

The head of the data will look like this:

	country_region	last_update	lat	lon	confirmed	deaths	recovered	active
0	Australia	2020-03-27 00:00:00	-25	133	3143	13	194	2936
1	Austria	2020-03-27 00:00:00	47.5162	14.5501	7317	58	225	7034
2	Canada	2020-03-27 00:00:00	60.001	-95.001	4046	40	184	0
3	China	2020-03-27 00:00:00	30.5928	114.305	81897	3296	74720	3881
4	Denmark	2020-03-27 00:00:00	56	10	2163	52	57	2054

Desired output

I would like to produce a plot that looks something like this:

Current output

Since I’ve been unable to reproduce the above plot, I hacked together a workaround that displays the information in a different way. My current output looks like this:

Current solution

The code that produces my plot is:

# Get data
import altair as alt
from vega_datasets import data

# Map plot
source = alt.topo_feature(data.world_110m.url, "countries")
base_map = (
    alt.Chart(source)
    .mark_geoshape(fill="white", stroke="gray")
    .properties(width=600, height=300)
    .project("naturalEarth1")
)

points = (
    alt.Chart(world_source)
    .mark_point()
    .encode(
        latitude="lat",
        longitude="lon",
        fill=alt.value("red"),
        size="confirmed:Q",
        stroke=alt.value(None),
     )
)

final_map = (
    (base_map + points)
    .configure_view(strokeWidth=0)
    .configure_mark(opacity=0.5,)
)

I’ve understood that using .transform_lookup() is one possible way to solve this, but all attempts at looking up the confirmed column in world_source have failed so far.

I would greatly appreciate any input!

[smu095]

Hi @jakevdp, thanks for the reply and useful tips.

You were right, it was just a matter of finding the appropriate ISO 3166-1 codes and performing a lookup on them. This is the (preliminary) result:

In case anyone is interested, this is the code to generate the plot above. It assumes that we have defined a lookup dataset (in my case world_source) with the required ISO 3166-1 codes.

source = alt.topo_feature(data.world_110m.url, "countries")

background = alt.Chart(source).mark_geoshape(fill="white")

foreground = (
    alt.Chart(source)
    .mark_geoshape(stroke="black", strokeWidth=0.15)
    .encode(
        color=alt.Color(
            "sick_per_100k:N", scale=alt.Scale(scheme="lightgreyred"), legend=None,
        ),
        tooltip=[
            alt.Tooltip("country_region:N", title="Country"),
            alt.Tooltip("sick_per_100k:Q", title="Cases pr. 100k"),
        ],
    )
    .transform_lookup(
        lookup="id",
        from_=alt.LookupData(world_source, "id", ["sick_per_100k", "country_region"]),
    )
)

final_map = (
    (background + foreground)
    .configure_view(strokeWidth=0)
    .properties(width=700, height=400)
    .project("naturalEarth1")
)

Thanks again @jakevdp, stay safe.

[reidjohnson]

@ThelmaSilva et al., here’s a full working example with the current data:

	country_region	last_update	lat	lon	confirmed	deaths	incident_rate	uid
0	Afghanistan	2021-12-17	33.939110	67.709953	157734	7332	405.190655	4
1	Albania	2021-12-17	41.153300	20.168300	205224	3158	7131.28084	8
2	Algeria	2021-12-17	28.033900	1.659600	213745	6171	487.434244	12
3	Andorra	2021-12-17	42.506300	1.521800	20549	134	26595.4831	20
4	Angola	2021-12-17	-11.202700	17.873900	65648	1737	199.742788	24

import altair as alt
import janitor
import pandas as pd
from vega_datasets import data

CASES_WORLDWIDE = "https://raw.githubusercontent.com/CSSEGISandData/COVID-19/web-data/data/cases_country.csv"


def get_worldwide_cases(url: str = CASES_WORLDWIDE):
    cases = pd.read_csv(url)
    cleaned = cases.clean_names(
    ).rename_column(
        "long_", "lon"
    ).transform_column(
        "last_update", lambda x: pd.to_datetime(x).normalize()
    )
    return cleaned


world_source = get_worldwide_cases()

source = alt.topo_feature(data.world_110m.url, "countries")

background = alt.Chart(source).mark_geoshape(fill="white")

foreground = (
    alt.Chart(source).mark_geoshape(
        stroke="black", strokeWidth=0.15
    ).encode(
        color=alt.Color(
            "incident_rate:N", scale=alt.Scale(scheme="lightgreyred"), legend=None,
        ),
        tooltip=[
            alt.Tooltip("country_region:N", title="Country"),
            alt.Tooltip("incident_rate:Q", title="Cases pr. 100k"),
        ],
    ).transform_lookup(
        lookup="id",
        from_=alt.LookupData(world_source, "uid", ["incident_rate", "country_region"]),
    )
)

chart = (
    (background + foreground)
    .configure_view(strokeWidth=0)
    .properties(width=700, height=400)
    .project("naturalEarth1")
)

chart