Flags

Get the flag emoji of a country using its 2 (or 3)-letter ISO country code

source

flag

 flag (cc)

*Get the flag emoji of a country using various identifiers.

Identifiers are case-insensitive so you don’t need to worry about whether it’s in lower or upper case:

  • Two-letter code: jp, mx, MX
  • Three-letter code: jpn, MEX
  • Country name: France, Jordan
  • Official name: THE REPUBLIC OF ARMENIA, The Federal Democratic Republic of Ethiopia
  • TLD (top-level domain): vn, br*
Type Details
cc str Country code, name, top-level domain or official country name
Returns str The Unicode emoji flag of the respective country

Get some country flags

flag("pt"), flag("prt")
('🇵🇹', '🇵🇹')
flag("jp"), flag("jpn"), flag("JPN"), flag("JAPAN"), flag("japan")
('🇯🇵', '🇯🇵', '🇯🇵', '🇯🇵', '🇯🇵')
flag("za")
'🇿🇦'

Use in a Jupyter notebook

The following html and md functions enable HTML and markdown output to contain dynamic values.

from IPython.display import display_html as html, display_markdown as md
from functools import partial

html = partial(html, raw=True)
md = partial(md, raw=True)

Use in a markdown string

md(f"# {flag('pr')}")

🇵🇷

Use in an HTML string for more granular control

html(
    f'<h1 style="font-size: 40pt; color: steelblue;">This is the flag of Sweden: <span style="font-size: 60pt;">{flag("se")}</span></h1>'
)

This is the flag of Sweden: 🇸🇪

Using flags in Plotly charts

population = pd.DataFrame(
    {
        "country": ["China", "India", "United States", "Indonesia", "Pakistan"],
        "population": [1439323776, 1380004385, 331002651, 273523615, 220892340],
        "code": ["cn", "in", "us", "id", "pk"],
    }
)
population
country population code
0 China 1439323776 cn
1 India 1380004385 in
2 United States 331002651 us
3 Indonesia 273523615 id
4 Pakistan 220892340 pk 
fig = px.bar(
    population,
    x="country",
    y="population",
    height=500,
    text=[flag(cc) for cc in population["code"]],
    template="ggplot2",
)
fig.data[0].textfont.size = 35
fig.data[0].marker.color = "steelblue"
fig
fig = px.bar(
    population.sort_values("population"),
    y="country",
    x="population",
    orientation="h",
    height=500,
    opacity=0.8,
    text=[flag(cc) for cc in population.sort_values("population")["code"]],
    template="ggplot2",
)
fig.data[0].textfont.size = 35
fig.data[0].marker.color = "teal"
fig
europe = gapminder().query("year==2007").query('continent=="Europe"')
europe["flag"] = [flag(cc) for cc in europe["iso_alpha"]]
europe.head()
country continent year lifeExp pop gdpPercap iso_alpha iso_num flag
23 Albania Europe 2007 76.423 3600523 5937.029526 ALB 8 🇦🇱
83 Austria Europe 2007 79.829 8199783 36126.492700 AUT 40 🇦🇹
119 Belgium Europe 2007 79.441 10392226 33692.605080 BEL 56 🇧🇪
155 Bosnia and Herzegovina Europe 2007 74.852 4552198 7446.298803 BIH 70 🇧🇦
191 Bulgaria Europe 2007 73.005 7322858 10680.792820 BGR 100 🇧🇬 
fig = px.scatter(
    europe,
    x="gdpPercap",
    y="lifeExp",
    height=600,
    text="flag",
    hover_data=["lifeExp", "pop", "gdpPercap"],
    hover_name=europe["flag"].add(" ").add(europe["country"]),
    title=f"Life expectancy ~ GDP per capita<br><b>Europe 2007</b>",
    template="plotly_white",
    log_y=False,
)
fig.data[0].textfont.size = 35
fig.data[0].hoverlabel.bgcolor = "white"
fig.data[0].hoverlabel.font.size = 20
fig

Cartograms improved

One misleading aspect of using maps to display data is that the area, shape, and relative position might give the wrong idea about the “size” of the metric for each country.

One way of dealing with this is creating “cartograms”, which are maps where the areas of countries are modified to reflect the size of the metric, while keeping the relative position in place.

They look very weird and even more misleading (some countries look like a tiny slice and can barely be visible due to the low value being visualized).

We can use flags to solve this issue, making diagrammatic cartograms. Flags have some interesting characteristics that can help in this:

  • They have the same size, so they won’t be misleading, and they are better than country names (where their lengths can also be misleading).
  • They are images, and therefore immediately recognizeable.

We will also make sure to display those flags in their respective locations, so it’s easy to find them. Now all we have to do is set text="flag" and set the font size to a multiple of our value that would fit in our chart. This needs a bit of trial and error, because of the variance in data values.

Code 
population = pd.read_html(
    "https://en.wikipedia.org/wiki/List_of_countries_and_dependencies_by_population"
)
pop = population[0][["Location", "Population"]].copy()
flags = []
for country in pop["Location"]:
    try:
        flags.append(flag(country))
    except Exception:
        flags.append("")
        continue

pop["flag"] = flags
pop = pop.dropna(subset=["flag"])

population[0][population[0]["Location"].str.contains("Czech")]

fig = px.scatter_geo(
    pop,
    locations="Location",
    locationmode="country names",
    text="flag",
    size="Population",
    hover_name="Location",
    title="Population",
    height=700,
    opacity=0,
)
fig.layout.geo.showframe = False
fig.layout.geo.lataxis.range = [-53, 76]
fig.layout.geo.lonaxis.range = [-125, 171.5]
fig.layout.geo.landcolor = "white"
fig.layout.geo.showcountries = False
fig.layout.geo.coastlinecolor = "white"
fig.data[0].textfont.size = np.sqrt(pop["Population"]).div(300).add(1).round()
fig.data[0].marker.color = "white"
fig