Skip to Content
AnalyticsSpatial Analysis & GISVisualization Guide

Visualization Guide

A library of styled, presentation-ready chart templates for health reporting, built with Matplotlib. Each entry is complete and self-contained: expand Show the code to see everything needed to reproduce the output below it — imports, data, styling, and layout. Swap the sample dataframe for your own and the styling carries over. The source notebook is analysis/spatial-analysis/Bar-Plots.ipynb.

The datasets in these examples are illustrative placeholders used only to demonstrate layout and styling. They contain no real health, patient, or facility data.

Horizontal bar

Horizontal bar with a highlighted category

Ranks named categories and draws the eye to one row with a single accent color, keeping the rest as neutral context.

Show the code

import matplotlib.pyplot as plt
import matplotlib.font_manager as fm
import pandas as pd
 
# Font setup
available = {f.name for f in fm.fontManager.ttflist}
FONT = next(
    (f for f in ["Gill Sans", "Gill Sans MT", "URW Classico", "Trebuchet MS"]
     if f in available),
    None
)
if FONT:
    plt.rcParams["font.family"] = FONT
    print(f"Using font: {FONT}")
else:
    print("Gill Sans not found – using matplotlib default. "
          "Install it and re-run for the exact look.")
 
# Data
HIGHLIGHT       = "Spain"
COLOR_DEFAULT   = "#1a4a7a"
COLOR_HIGHLIGHT = "#5fc4c0"
BG_COLOR        = "#dce8f0"
 
df = pd.DataFrame({
    "country": ["Germany", "Spain", "Austria", "France", "Luxembourg",
                "Portugal", "Britain", "Denmark", "Sweden"],
    "consultations_per_person": [8.2, 7.5, 6.8, 6.7, 6.3, 5.8, 5.1, 4.7, 2.9],
})
df["highlight"] = df["country"] == HIGHLIGHT
df["color"]     = df["highlight"].map({True: COLOR_HIGHLIGHT, False: COLOR_DEFAULT})
df["y_pos"]     = range(len(df) - 1, -1, -1)   # top-to-bottom order
 
print(df.to_string(index=False))
 
fig, ax = plt.subplots(figsize=(7, 5))
fig.patch.set_facecolor(BG_COLOR)
ax.set_facecolor(BG_COLOR)
 
# Bars
ax.barh(df["y_pos"], df["consultations_per_person"],
        color=df["color"], height=0.6)
 
# Axes
ax.set_xlim(0, 8.5)
ax.set_xticks([0, 2, 4, 6, 8])
ax.xaxis.set_tick_params(labelsize=9, colors="#555555")
ax.xaxis.tick_top()
ax.set_yticks(df["y_pos"])
ax.set_yticklabels(
    [f"$\\bf{{{c}}}$" if c == HIGHLIGHT else c for c in df["country"]],
    fontsize=10
)
ax.tick_params(axis="y", length=0)
 
# Grid & spines
ax.xaxis.grid(True, color="white", linewidth=0.8, zorder=0)
ax.set_axisbelow(True)
for spine in ax.spines.values():
    spine.set_visible(False)
ax.tick_params(axis="x", length=0)
 
# Red badge
fig.add_artist(plt.Rectangle((0.02, 0.975), 0.055, 0.016,
                              transform=fig.transFigure,
                              color="#c0392b", clip_on=False))
 
# Title block
fig.text(0.02, 0.965, "Salud", fontsize=12, fontweight="bold",
         va="top", color="#1a1a1a", transform=fig.transFigure)
fig.text(0.02, 0.930, "Doctors' consultations per person", fontsize=9,
         va="top", color="#333333", transform=fig.transFigure)
fig.text(0.02, 0.905, "Selected countries, 2009", fontsize=8.5,
         va="top", color="#555555", transform=fig.transFigure)
 
# Footer
fig.text(0.02, 0.02, "Source: Health Management Information System",
         fontsize=8, color="#666666", transform=fig.transFigure)
fig.text(0.79, 0.02, "Health Intelligence Center",
         fontsize=8, color="#666666", transform=fig.transFigure)
 
# rect=[left, bottom, right, top] — lower 'top' pushes the plot area down
plt.tight_layout(rect=[0, 0.03, 1, 0.78])
plt.show()
charts/doctors_consultations.png

Horizontal bar, two series

Compares two measures per category (for example two periods or two indicators) with paired bars.

Show the code

import matplotlib.pyplot as plt
import matplotlib.font_manager as fm
import pandas as pd
import numpy as np
 
# Font setup
available = {f.name for f in fm.fontManager.ttflist}
FONT = next(
    (f for f in ["Gill Sans", "Gill Sans MT", "URW Classico", "Trebuchet MS"]
     if f in available),
    None
)
if FONT:
    plt.rcParams["font.family"] = FONT
    print(f"Using font: {FONT}")
else:
    print("Gill Sans not found – using matplotlib default.")
 
# Data
df = pd.DataFrame({
    "country":      ["Rubavu", "Rusizi", "Bugesera", "Nyabihu", "Musanze",
                     "Burera", "Rulindo", "Gasabo", "Nyamasheke"],
    "middle_class": [80, 72, 60, 58, 55, 50, 62, 50, 48],
    "low_income":   [68, 63, 50, 62, 38, 42, 55, 48, 42],
})
 
# Reverse so Chile appears on top
df = df.iloc[::-1].reset_index(drop=True)
 
print(df.to_string(index=False))
 
# Colors
COLOR_MIDDLE  = "#1a4a7a"
COLOR_LOW     = "#5fc4c0"
BG_COLOR      = "#dce8f0"
 
# Layout
n       = len(df)
y       = np.arange(n)
height  = 0.35          # bar height for each trace
 
fig, ax = plt.subplots(figsize=(7, 5.5))
fig.patch.set_facecolor(BG_COLOR)
ax.set_facecolor(BG_COLOR)
 
# Bars
ax.barh(y + height / 2, df["middle_class"], height=height,
        color=COLOR_MIDDLE, label="Middle class")
ax.barh(y - height / 2, df["low_income"],  height=height,
        color=COLOR_LOW,    label="Low income")
 
# Axes
ax.set_xlim(0, 85)
ax.set_xticks([0, 20, 40, 60, 80])
ax.xaxis.set_tick_params(labelsize=9, colors="#555555")
ax.xaxis.tick_top()
ax.set_yticks(y)
ax.set_yticklabels(df["country"], fontsize=10, fontweight="bold")
ax.tick_params(axis="y", length=0)
 
# Grid & spines
ax.xaxis.grid(True, color="white", linewidth=0.8, zorder=0)
ax.set_axisbelow(True)
for spine in ax.spines.values():
    spine.set_visible(False)
ax.tick_params(axis="x", length=0)
 
# Legend
legend = ax.legend(
    loc="lower right",
    fontsize=9,
    frameon=False,
    handlelength=1.2,
    handleheight=0.9,
    handletextpad=0.5,
)
 
# Red badge
fig.add_artist(plt.Rectangle((0.02, 0.965), 0.045, 0.016,
                              transform=fig.transFigure,
                              color="#c0392b", clip_on=False))
 
# Title block
fig.text(0.02, 0.955, "Defenders of democracy",
         fontsize=12, fontweight="bold", va="top",
         color="#1a1a1a", transform=fig.transFigure)
fig.text(0.02, 0.915,
         "Respondents saying that honest elections held regularly\n"
         "with a choice of at least two political parties are very\n"
         "important, 2007, %",
         fontsize=8.5, va="top", color="#444444",
         transform=fig.transFigure, linespacing=1.5)
 
# Footer
fig.text(0.02, 0.02, "Source: Pew Global Attitudes Survey",
         fontsize=8, color="#666666", transform=fig.transFigure)
fig.text(0.72, 0.02, "Health Intelligence Center",
         fontsize=8, color="#666666", transform=fig.transFigure)
 
plt.tight_layout(rect=[0, 0.03, 1, 0.78])
plt.show()
charts/defenders_of_democracy.png

Grouped horizontal bar with sub-categories

Breaks each category into grouped sub-category bars with group labels.

Show the code

import matplotlib.pyplot as plt
import matplotlib.font_manager as fm
import matplotlib.text as mtext
import pandas as pd
import numpy as np
 
# Font setup
available = {f.name for f in fm.fontManager.ttflist}
FONT = next(
    (f for f in ["Gill Sans", "Gill Sans MT", "URW Classico", "Trebuchet MS"]
     if f in available),
    None
)
if FONT:
    plt.rcParams["font.family"] = FONT
    print(f"Using font: {FONT}")
else:
    print("Gill Sans not found – using matplotlib default.")
 
# Data
df = pd.DataFrame({
    "country":   ["Burera", "Nyabihu", "Rulindo",    "Gasabo",
                  "Musanze","Rwamagana","Rubavu",    "Gakenke",
                  "Nyarugenge",   "Gicumbi"],
    "year":      [2011, 2012] * 5,
    "june":      [3.0, 1.8, 2.1, 1.4, 2.0, 1.5, 0.8, 1.1, 1.0, 0.3],
    "september": [2.7, 1.3, 1.7, 1.0, 1.6, 1.1, 0.7, 0.3, 0.6, 0.1],
})
 
print(df.to_string(index=False))
 
# Colors
COLOR_JUNE = "#1a4a7a"
COLOR_SEP  = "#5fc4c0"
BG_COLOR   = "#dce8f0"
 
# Build y positions
# Layout (bottom→top):  Italy_2012, Italy_2011, gap, Spain_2012, Spain_2011 …
# Each year-row occupies 2 bar slots stacked: june on top, sep below.
# june centre  = y_row + BAR_H/2
# sep  centre  = y_row - BAR_H/2
BAR_H      = 0.30   # height of each individual bar
ROW_STEP   = BAR_H * 2 + 0.05   # vertical space per year-row
GROUP_GAP  = 0.30               # extra gap between country groups
 
countries = df["country"].unique()   # original order: Germany … Italy
 
positions = []
y = 0
# Build bottom-up: last country first, within each country 2012 first
for country in reversed(countries):
    sub = df[df["country"] == country].sort_values("year")  # 2011, 2012
    # Place 2012 at bottom of group, 2011 above it
    for _, row in sub.iterrows():
        positions.append({
            "country": country,
            "year": int(row["year"]),
            "june": row["june"],
            "sep":  row["september"],
            "y":    y
        })
        y += ROW_STEP
    y += GROUP_GAP
 
pos_df = pd.DataFrame(positions)
# After building bottom-up: lower y = 2012, higher y = 2011 ✓
 
# Plot
fig, ax = plt.subplots(figsize=(6, 6.5))
fig.patch.set_facecolor(BG_COLOR)
ax.set_facecolor(BG_COLOR)
 
for _, r in pos_df.iterrows():
    # June on top (+BAR_H/2), September below (-BAR_H/2)
    ax.barh(r["y"] + BAR_H / 2, r["june"], height=BAR_H,
            color=COLOR_JUNE, align="center")
    ax.barh(r["y"] - BAR_H / 2, r["sep"],  height=BAR_H,
            color=COLOR_SEP,  align="center")
 
# Y-axis: year labels only, country drawn via ax.text
ax.set_yticks(pos_df["y"])
ax.set_yticklabels([str(r["year"]) for _, r in pos_df.iterrows()],
                   fontsize=9, color="#555555")
ax.tick_params(axis="y", length=0, pad=2)
 
# Draw country name centred between its two year rows
for country in countries:
    rows = pos_df[pos_df["country"] == country]
    mid_y = rows["y"].mean()
    ax.text(-0.18, mid_y, country,
            ha="right", va="center", fontsize=9, fontweight="bold",
            transform=ax.get_yaxis_transform())
 
# X-axis
ax.set_xlim(0, 3.3)
ax.set_xticks([0, 1, 2, 3])
ax.xaxis.set_tick_params(labelsize=9, colors="#555555")
ax.xaxis.tick_top()
ax.tick_params(axis="x", length=0)
 
# Grid & spines
ax.xaxis.grid(True, color="white", linewidth=0.8, zorder=0)
ax.set_axisbelow(True)
for spine in ax.spines.values():
    spine.set_visible(False)
 
# Legend
from matplotlib.patches import Patch
legend_elements = [Patch(facecolor=COLOR_JUNE, label="June"),
                   Patch(facecolor=COLOR_SEP,  label="September")]
ax.legend(handles=legend_elements, loc="lower right", fontsize=8,
          frameon=False, title="Forecasts made in:", title_fontsize=8)
 
# Red badge
fig.add_artist(plt.Rectangle((0.02, 0.975), 0.045, 0.013,
                              transform=fig.transFigure,
                              color="#c0392b", clip_on=False))
 
# Title block
fig.text(0.02, 0.968, "Autumn fall", fontsize=12, fontweight="bold",
         va="top", color="#1a1a1a", transform=fig.transFigure)
fig.text(0.02, 0.936, "GDP, % increase on previous year", fontsize=9,
         va="top", color="#444444", transform=fig.transFigure)
 
# Footer
fig.text(0.02, 0.02, "Source: HMIS",
         fontsize=8, color="#666666", transform=fig.transFigure)
fig.text(0.79, 0.02, "Health Intelligence Center",
         fontsize=8, color="#666666", transform=fig.transFigure)
 
plt.tight_layout(rect=[0, 0.03, 1, 0.80])
plt.show()
charts/autumn_fall_gdp.png

Ranked horizontal bar with emphasis

A longer ranked list where one bar carries the story.

Show the code

import matplotlib.pyplot as plt
import matplotlib.font_manager as fm
import pandas as pd
import numpy as np
 
# Font setup
available = {f.name for f in fm.fontManager.ttflist}
FONT = next(
    (f for f in ["Gill Sans", "Gill Sans MT", "URW Classico", "Trebuchet MS"]
     if f in available),
    None
)
if FONT:
    plt.rcParams["font.family"] = FONT
    print(f"Using font: {FONT}")
else:
    print("Gill Sans not found – using matplotlib default.")
 
# Data
df = pd.DataFrame({
    "sector": ["Internet", "Software", "Tech hardware", "IT services",
               "Health care", "Total", "Leisure", "Chemicals",
               "Oil, gas and fuels", "Metals and mining", "Property",
               "Energy equipment"],
    "change": [57, 34, 20, 16, 10, 4, -3, -5, -18, -22, -28, -58],
    "bold":   [False, False, False, False, False, True,
               False, False, False, False, False, False],
})
 
# Reverse so Internet is on top
df = df.iloc[::-1].reset_index(drop=True)
 
print(df.to_string(index=False))
 
# Colors
COLOR_BAR = "#1a4a7a"
COLOR_ZERO = "#c0392b"
BG_COLOR   = "#dce8f0"
 
# Plot
fig, ax = plt.subplots(figsize=(6, 6.5))
fig.patch.set_facecolor(BG_COLOR)
ax.set_facecolor(BG_COLOR)
 
y = np.arange(len(df))
ax.barh(y, df["change"], color=COLOR_BAR, height=0.6, zorder=2)
 
# Zero line (red)
ax.axvline(x=0, color=COLOR_ZERO, linewidth=1.5, zorder=3)
 
# Axes
ax.set_xlim(-65, 65)
ax.set_xticks([-60, -40, -20, 0, 20, 40, 60])
ax.xaxis.set_tick_params(labelsize=9, colors="#555555")
ax.xaxis.tick_top()
ax.tick_params(axis="x", length=0)
 
ax.set_yticks(y)
ax.set_yticklabels(df["sector"], fontsize=10)
ax.tick_params(axis="y", length=0)
 
# Bold "Total" label
for label, (_, row) in zip(ax.get_yticklabels(), df.iterrows()):
    if row["bold"]:
        label.set_fontweight("bold")
 
# Grid & spines
ax.xaxis.grid(True, color="white", linewidth=0.8, zorder=0)
ax.set_axisbelow(True)
for spine in ax.spines.values():
    spine.set_visible(False)
 
# Red badge
fig.add_artist(plt.Rectangle((0.02, 0.975), 0.045, 0.013,
                              transform=fig.transFigure,
                              color="#c0392b", clip_on=False))
 
# Title block
fig.text(0.02, 0.968, "Another dotcom boom", fontsize=12, fontweight="bold",
         va="top", color="#1a1a1a", transform=fig.transFigure)
fig.text(0.02, 0.935, "Change in worldwide capital spending*, %",
         fontsize=9, va="top", color="#444444", transform=fig.transFigure)
fig.text(0.02, 0.912, "2014-17 forecast",
         fontsize=9, va="top", color="#444444", transform=fig.transFigure)
 
# Footer
fig.text(0.02, 0.02, "Source: Goldman Sachs",
         fontsize=8, color="#666666", transform=fig.transFigure)
fig.text(0.68, 0.02, "*In dollar terms",
         fontsize=8, color="#666666", transform=fig.transFigure)
 
plt.tight_layout(rect=[0, 0.03, 1, 0.78])
plt.show()
charts/dotcom_boom_capex.png

Vertical column

Vertical column chart

Ordered categories or a short time series reading left to right.

Show the code

import matplotlib.pyplot as plt
import matplotlib.font_manager as fm
import pandas as pd
 
# Font setup
available = {f.name for f in fm.fontManager.ttflist}
FONT = next(
    (f for f in ["Gill Sans", "Gill Sans MT", "URW Classico", "Trebuchet MS"]
     if f in available),
    None
)
if FONT:
    plt.rcParams["font.family"] = FONT
    print(f"Using font: {FONT}")
else:
    print("Gill Sans not found – using matplotlib default.")
 
# Data
df = pd.DataFrame({
    "year":     ["2004","06","08","10","11","12","13","14","15*"],
    "gdp":      [ 11.7,  12.6, 11.2, 10.0, 11.2, 10.9,  9.7, 10.3,  8.7],
    "estimate": [ False, False, False, False, False, False, False, False, True],
})
 
# Colors
COLOR_DEFAULT  = "#1a4a7a"
COLOR_ESTIMATE = "#5fc4c0"
BG_COLOR       = "#dce8f0"
 
df["color"] = df["estimate"].map({False: COLOR_DEFAULT, True: COLOR_ESTIMATE})
 
print(df.to_string(index=False))
 
# Plot
fig, ax = plt.subplots(figsize=(5, 5))
fig.patch.set_facecolor(BG_COLOR)
ax.set_facecolor(BG_COLOR)
 
x = range(len(df))
ax.bar(x, df["gdp"], color=df["color"], width=0.75, zorder=2)
 
# Axes
ax.set_xlim(-0.5, len(df) - 0.5)
ax.set_ylim(0, 14)
ax.set_yticks([0, 2, 4, 6, 8, 10, 12, 14])
ax.yaxis.set_tick_params(labelsize=9, colors="#555555")
ax.yaxis.tick_right()                        # y-axis on the right
ax.tick_params(axis="y", length=0)
 
ax.set_xticks(list(x))
ax.set_xticklabels(df["year"], fontsize=9, color="#555555")
ax.tick_params(axis="x", length=0)
 
# Grid & spines
ax.yaxis.grid(True, color="white", linewidth=0.8, zorder=0)
ax.set_axisbelow(True)
for spine in ax.spines.values():
    spine.set_visible(False)
 
# Red badge
fig.add_artist(plt.Rectangle((0.02, 0.975), 0.055, 0.014,
                              transform=fig.transFigure,
                              color="#c0392b", clip_on=False))
 
# Title block
fig.text(0.02, 0.968, "African lion", fontsize=12, fontweight="bold",
         va="top", color="#1a1a1a", transform=fig.transFigure)
fig.text(0.02, 0.935, "Ethiopia's GDP, % change on a year earlier",
         fontsize=9, va="top", color="#444444", transform=fig.transFigure)
 
# Footer
fig.text(0.02, 0.02, "Source: IMF",
         fontsize=8, color="#666666", transform=fig.transFigure)
fig.text(0.02, 0.055, "*Estimate",
         fontsize=8, color="#666666", transform=fig.transFigure)
fig.text(0.79, 0.02, "Health Intelligence Center",
         fontsize=8, color="#666666", transform=fig.transFigure)
 
plt.tight_layout(rect=[0, 0.06, 1, 0.80])
plt.show()
charts/african_lion_gdp.png

Vertical column chart, alternate styling

The same column pattern with a different palette and emphasis.

Show the code

import matplotlib.pyplot as plt
import matplotlib.font_manager as fm
import matplotlib.patches as mpatches
import pandas as pd
import numpy as np
 
# Font setup
available = {f.name for f in fm.fontManager.ttflist}
FONT = next(
    (f for f in ["Gill Sans", "Gill Sans MT", "URW Classico", "Trebuchet MS"]
     if f in available),
    None
)
if FONT:
    plt.rcParams["font.family"] = FONT
    print(f"Using font: {FONT}")
else:
    print("Gill Sans not found – using matplotlib default.")
 
# Data
df = pd.DataFrame({
    "period":       ["20 years", "10 years", "5 years"],
    "equities":     [8.5,  7.5,  8.5],
    "all_property": [9.5,  8.0,  13.5],
    "rural":        [14.5, 13.5, 13.0],
    "forestry":     [11.0, 18.5, 20.5],
})
 
print(df.to_string(index=False))
 
# Colors
COLORS = {
    "equities":     "#1a4a7a",
    "all_property": "#5fc4c0",
    "rural":        "#c9a84c",
    "forestry":     "#2d7a4a",
}
LABELS = {
    "equities":     "Equities",
    "all_property": "All property",
    "rural":        "Rural property",
    "forestry":     "Forestry",
}
BG_COLOR = "#dce8f0"
 
keys = list(COLORS.keys())
n_groups = len(df)
n_bars   = len(keys)
width    = 0.18          # width of each bar
group_w  = width * n_bars + 0.08   # total group width incl. gap
 
x = np.arange(n_groups)
 
# Plot
fig, ax = plt.subplots(figsize=(6, 5))
fig.patch.set_facecolor(BG_COLOR)
ax.set_facecolor(BG_COLOR)
 
for i, key in enumerate(keys):
    offsets = x - (n_bars - 1) * width / 2 + i * width
    ax.bar(offsets, df[key], width=width, color=COLORS[key],
           label=LABELS[key], zorder=2)
 
# Axes
ax.set_xlim(-0.5, n_groups - 0.5)
ax.set_ylim(0, 22)
ax.set_yticks([0, 5, 10, 15, 20])
ax.yaxis.set_tick_params(labelsize=9, colors="#555555")
ax.yaxis.tick_right()
ax.tick_params(axis="y", length=0)
 
ax.set_xticks(x)
ax.set_xticklabels(df["period"], fontsize=9, color="#555555")
ax.tick_params(axis="x", length=0)
 
# Grid & spines
ax.yaxis.grid(True, color="white", linewidth=0.8, zorder=0)
ax.set_axisbelow(True)
for spine in ax.spines.values():
    spine.set_visible(False)
 
# Legend
handles = [mpatches.Patch(facecolor=COLORS[k], label=LABELS[k]) for k in keys]
ax.legend(handles=handles, ncol=2, fontsize=8, frameon=False,
          loc="upper left", bbox_to_anchor=(0, 1.02),
          handlelength=1.0, handleheight=0.9, columnspacing=1.0)
 
# Red badge
fig.add_artist(plt.Rectangle((0.02, 0.975), 0.055, 0.014,
                              transform=fig.transFigure,
                              color="#c0392b", clip_on=False))
 
# Title block
fig.text(0.02, 0.968, "The mighty jungle", fontsize=12, fontweight="bold",
         va="top", color="#1a1a1a", transform=fig.transFigure)
fig.text(0.02, 0.935, "Britain, annualised rate of return\non investment*, %",
         fontsize=9, va="top", color="#444444",
         transform=fig.transFigure, linespacing=1.5)
 
# Footer
fig.text(0.02, 0.02, "Sources: MSCI; JP Morgan",
         fontsize=8, color="#666666", transform=fig.transFigure)
fig.text(0.75, 0.02, "*To 2014",
         fontsize=8, color="#666666", transform=fig.transFigure)
fig.text(0.79, 0.055, "Health Intelligence Center",
         fontsize=8, color="#666666", transform=fig.transFigure)
 
plt.tight_layout(rect=[0, 0.06, 1, 0.75])
plt.show()
charts/mighty_jungle_returns.png

Vertical column chart with a highlighted bar

A column chart that highlights a single period or category.

Show the code

import matplotlib.pyplot as plt
import matplotlib.font_manager as fm
import matplotlib.patches as mpatches
import numpy as np
 
# Font setup
available = {f.name for f in fm.fontManager.ttflist}
FONT = next(
    (f for f in ["Gill Sans", "Gill Sans MT", "URW Classico", "Trebuchet MS"]
     if f in available),
    None
)
if FONT:
    plt.rcParams["font.family"] = FONT
    print(f"Using font: {FONT}")
else:
    print("Gill Sans not found – using matplotlib default.")
 
# Data
# Approximate net private capital inflows/outflows in $bn, 1994–2011
# Values estimated from the chart image
years = [
    1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003,
    2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011
]
values = [
    -4, -5, -23, -18, -8, -18, -25, -15, -8, -7,
    -35, -30, 40, 80, -130, -50, -35, 5
]
 
# Two-tone: teal for positive inflows, darker teal for outflows (negative)
# The chart uses a medium teal for most bars and a lighter teal for the
# 2006-2008 positive bars — we'll use a single accent for positive values
COLOR_NEG  = "#3a9aa0"   # muted teal (outflows)
COLOR_POS  = "#1a6a7a"   # darker teal (inflows)
COLOR_ZERO = "#c0392b"   # red zero line
BG_COLOR   = "#dce8f0"   # light blue-grey Economist background
 
bar_colors = [COLOR_POS if v >= 0 else COLOR_NEG for v in values]
 
# Plot
fig, ax = plt.subplots(figsize=(5, 5.5))
fig.patch.set_facecolor(BG_COLOR)
ax.set_facecolor(BG_COLOR)
 
x = np.arange(len(years))
ax.bar(x, values, color=bar_colors, width=0.75, zorder=2)
 
# Zero / baseline (red)
ax.axhline(y=0, color=COLOR_ZERO, linewidth=1.5, zorder=3)
 
# Axes
ax.set_ylim(-160, 120)
ax.set_yticks([100, 50, 0, -50, -100, -150])
ax.yaxis.set_tick_params(labelsize=9, colors="#555555")
ax.yaxis.tick_right()
ax.tick_params(axis="y", length=0)
 
# X-axis: show subset of year labels to match original
label_map = {
    1994: "1994", 1996: "96", 1998: "98", 2000: "2000", 2002: "02",
    2004: "04",   2006: "06", 2008: "08", 2011: "11*"
}
xtick_positions = [i for i, yr in enumerate(years) if yr in label_map]
xtick_labels    = [label_map[years[i]] for i in xtick_positions]
 
ax.set_xticks(xtick_positions)
ax.set_xticklabels(xtick_labels, fontsize=8.5, color="#555555")
ax.tick_params(axis="x", length=0)
ax.xaxis.tick_bottom()
 
# Grid & spines
ax.yaxis.grid(True, color="white", linewidth=0.8, zorder=0)
ax.set_axisbelow(True)
for spine in ax.spines.values():
    spine.set_visible(False)
 
# Red badge (Economist house style)
fig.add_artist(plt.Rectangle(
    (0.02, 0.975), 0.045, 0.013,
    transform=fig.transFigure,
    color="#c0392b", clip_on=False
))
 
# Title block
fig.text(0.02, 0.968, "The money drain",
         fontsize=12, fontweight="bold",
         va="top", color="#1a1a1a", transform=fig.transFigure)
fig.text(0.02, 0.935, "Net private capital inflows/outflows, $bn",
         fontsize=9, va="top", color="#444444", transform=fig.transFigure)
 
# Footer
fig.text(0.02, 0.02, "Source: Central Bank of Russia",
         fontsize=8, color="#666666", transform=fig.transFigure)
fig.text(0.55, 0.02, "*To June 30th",
         fontsize=8, color="#666666", transform=fig.transFigure)
 
plt.tight_layout(rect=[0, 0.04, 1, 0.80])
plt.show()
charts/money_drain.png

Scatter

Annotated scatter plot

Relates two continuous variables, labelling points directly instead of using a legend.

Show the code

import matplotlib
matplotlib.rcParams['text.usetex'] = False
import matplotlib.pyplot as plt
import matplotlib.font_manager as fm
import matplotlib.patches as mpatches
import numpy as np
 
available = {f.name for f in fm.fontManager.ttflist}
FONT = next((f for f in ["Gill Sans","Gill Sans MT","URW Classico","Trebuchet MS"] if f in available), None)
if FONT:
    plt.rcParams["font.family"] = FONT
 
C_MW    = "#8b3030"   # Midwest   — dark red (highlighted)
C_OTHER = "#3bbfbf"   # All other — teal
C_GRAY  = "#9e9e9e"   # Large cities — grey
BG      = "#dce8f0"
RED_LINE= "#c0392b"
RED_BAR = "#c0392b"
 
np.random.seed(42)
 
def make_counties(n, health_mu, health_sd, swing_mu, swing_sd, size_mu, size_sd):
    """Simulate county-level data. Replace with real arrays when available."""
    health = np.random.normal(health_mu, health_sd, n)
    swing  = np.random.normal(swing_mu,  swing_sd,  n)
    size   = np.abs(np.random.normal(size_mu, size_sd, n)) + 10
    return health, swing, size
 
# Other regions (teal) — spread across full x-range, mixed swing
h_ot, s_ot, sz_ot = make_counties(600, 48, 8,  -2, 16, 20, 18)
 
# Midwest (dark red) — clustered at lower health, stronger R swing
h_mw, s_mw, sz_mw = make_counties(450, 38, 5,  16, 10, 18, 14)
 
# Large cities (grey) — few, larger bubbles, less partisan swing
h_gr  = np.array([35,   45,   52,   58,   40,   30  ])
s_gr  = np.array([-5,   -8,  -12,   -2,  -15,   -3  ])
sz_gr = np.array([180,  220,  300,  250,  160,  140  ])
 
# Named/highlighted counties — white-outlined dots with leader lines
named = {
    "Jefferson, OH": (32, 42),
    "Knox, OH":      (43, 32),
}
 
# Figure
fig, ax = plt.subplots(figsize=(7.4, 5.6))
fig.patch.set_facecolor(BG)
ax.set_facecolor(BG)
fig.subplots_adjust(left=0.11, right=0.76, top=0.77, bottom=0.18)
 
# Layer order: other (bottom) → grey cities → Midwest (top)
ax.scatter(h_ot, s_ot, s=sz_ot, color=C_OTHER, alpha=0.55,
           linewidths=0, zorder=2)
ax.scatter(h_gr, s_gr, s=sz_gr, color=C_GRAY, alpha=0.65,
           linewidths=0.5, edgecolors='white', zorder=3)
ax.scatter(h_mw, s_mw, s=sz_mw, color=C_MW, alpha=0.60,
           linewidths=0, zorder=4)
 
# Named counties — white-filled dot with dark outline + leader line + label
for name, (x, y) in named.items():
    ax.scatter(x, y, s=38, color='white', linewidths=1.2,
               edgecolors=C_MW, zorder=6)
    ox = 1.2 if "Knox" in name else 1.0
    ax.annotate(name, xy=(x, y), xytext=(x + ox, y + 3.5),
                fontsize=8, color="#1a1a1a", zorder=7,
                arrowprops=dict(arrowstyle="-", color="#555555", lw=0.7))
 
# Region label inside cluster
ax.text(37, 14, "Midwest", fontsize=13, fontweight="bold",
        color="black", alpha=0.85, zorder=5, style='italic')
 
# Zero reference line
ax.axhline(0, color=RED_LINE, linewidth=1.0, zorder=2)
 
# Axes
ax.set_xlim(18, 72)
ax.set_ylim(-55, 55)
ax.set_xticks([20, 30, 40, 50, 60, 70])
ax.set_yticks([-50, -25, 0, 25, 50])
ax.xaxis.set_tick_params(labelsize=8.5, colors="#555555", length=0)
ax.yaxis.set_tick_params(labelsize=8.5, colors="#555555", length=0)
ax.xaxis.grid(True, color="white", linewidth=0.8, zorder=0)
ax.yaxis.grid(True, color="white", linewidth=0.8, zorder=0)
ax.set_axisbelow(True)
for sp in ax.spines.values():
    sp.set_visible(False)
 
# Axis labels
ax.set_xlabel("Index of county health metrics*", fontsize=9,
              color="#444444", labelpad=22)
ax.text(0.02, -0.13, "◄ WORSE HEALTH", fontsize=7.5, color="#555555",
        transform=ax.transAxes, va='top')
ax.text(0.98, -0.13, "BETTER HEALTH ►", fontsize=7.5, color="#555555",
        ha='right', transform=ax.transAxes, va='top')
ax.text(-0.10, 0.75, "MORE REPUBLICAN", fontsize=7.5, color="#555555",
        transform=ax.transAxes, va='center', ha='center', rotation=90)
ax.text(-0.10, 0.25, "LESS REPUBLICAN", fontsize=7.5, color="#555555",
        transform=ax.transAxes, va='center', ha='center', rotation=90)
 
# Right-side y-axis description
fig.text(0.77, 0.77,
         "Change in Republican margin over\nDemocrats, 2012-16, % points",
         fontsize=7.8, color="#444444", va="top", ha="left",
         linespacing=1.4, transform=fig.transFigure)
 
# Bubble size legend
lx, ly = 0.790, 0.62
circ = mpatches.Circle((lx + 0.025, ly - 0.018), 0.026,
    transform=fig.transFigure, fill=False,
    edgecolor="#555555", linewidth=0.9, clip_on=False)
fig.add_artist(circ)
fig.text(lx + 0.060, ly, "Voting\neligible\npopulation",
    fontsize=7.5, color="#1a1a1a", va="center",
    transform=fig.transFigure, linespacing=1.3)
 
# Red badge + Title
fig.add_artist(plt.Rectangle((0.06, 0.942), 0.038, 0.008,
    transform=fig.transFigure, color=RED_BAR, clip_on=False))
fig.text(0.06, 0.938, "Vitality and the vote",
    fontsize=11, fontweight="bold", va="top", color="#1a1a1a",
    transform=fig.transFigure)
fig.text(0.06, 0.910,
    "United States, health metrics against swing to Donald Trump",
    fontsize=8.5, va="top", color="#444444", transform=fig.transFigure)
fig.text(0.06, 0.890, "By county",
    fontsize=8, va="top", color="#444444", transform=fig.transFigure)
 
# Footer
fig.text(0.06, 0.022,
    "Sources: Atlas of US Presidential Elections; Census Bureau;\n"
    "IPUMS, University of Minnesota; Institute for Health Metrics\n"
    "and Evaluation; The Economist",
    fontsize=7, color="#666666", transform=fig.transFigure, linespacing=1.4)
fig.text(0.52, 0.022,
    "*Weighted index of obesity, diabetes,\nheavy drinking, physical exercise and\nlife expectancy, 2010-12",
    fontsize=7, color="#666666", transform=fig.transFigure, linespacing=1.4)
 
print("Saved: vitality_vote_midwest.png")
charts/vitality_vote_midwest.png
Last updated on
Climate MapsData Reference