vizop

Opinionated data visualization for Python. Publication-quality charts with minimal configuration.

vizop produces clean, presentation-ready charts inspired by the visual style of the NYT and FiveThirtyEight — left-aligned titles, no chartjunk, smart defaults. Built on matplotlib, driven by DataFrames.

---

Gallery

Line	Bar	Scatter	Slope
Waffle	Raincloud	Parliament	Bump

---

Installation

pip install vizop

or with uv:

uv add vizop

---

Quick Start

import pandas as pd
import vizop

df = pd.DataFrame({
    "month": ["Jan", "Feb", "Mar", "Apr", "May", "Jun",
              "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"],
    "visitors": [1200, 1350, 1800, 2400, 2800, 3200,
                 3500, 3300, 2900, 2100, 1600, 1400],
})

chart = vizop.line(
    df, x="month", y="visitors",
    title="Monthly Website Visitors",
    subtitle="Pageviews peaked in July before seasonal decline",
    source="Internal analytics",
)
chart.show()

Every chart function returns a Chart object — call .show() to display, .save("file.png") to export, or just let it render inline in Jupyter.

---

Charts

Line Chart

Basic single-series line:

df = pd.DataFrame({
    "month": ["Jan", "Feb", "Mar", "Apr", "May", "Jun",
              "Jul", "Aug", "Sep", "Oct", "Nov", "Dec"],
    "temperature": [32, 35, 45, 55, 65, 75, 82, 80, 70, 58, 45, 35],
})

vizop.line(
    df, x="month", y="temperature",
    title="Average Monthly Temperature",
    subtitle="Temperatures in degrees Fahrenheit, New York City",
    source="National Weather Service",
)

Multi-series with highlight — the highlighted series stays vivid while others are muted. Labels appear at line endpoints instead of a legend box:

vizop.line(
    df, x="month", y="ridership", group="mode",
    highlight="Subway",
    title="NYC Transit Ridership by Mode",
    subtitle="Daily riders in thousands — subway dominates year-round",
    source="MTA Open Data",
)

Key Parameters

Parameter	Type	Description
x	str	Column for the x-axis
y	str | list[str]	Column(s) for y-axis. List for wide-format multi-series
group	str	Column to group by (long format)
highlight	str | list[str]	Series to highlight; others are muted
show_area	bool	Fill area under the line (single-series)
zero_baseline	bool	Force y-axis to start at 0
show_last_value	bool	Display value at the last data point
highlight_range	tuple	(start, end) x-values to shade

---

Bar Chart

Bars are horizontal and sorted descending by default — the most important value is always at the top:

df = pd.DataFrame({
    "language": ["Python", "JavaScript", "Java", "C++", "TypeScript",
                 "Go", "Rust", "Swift"],
    "satisfaction": [73, 61, 42, 48, 78, 80, 87, 65],
})

vizop.bar(
    df, x="language", y="satisfaction",
    title="Developer Satisfaction by Language",
    subtitle="Percent of developers who would use the language again",
    source="2025 Developer Survey",
)

Grouped bars for comparing across categories:

vizop.bar(
    df, x="region", y="revenue", group="year",
    orientation="vertical", mode="grouped",
    title="Regional Revenue Growth",
    subtitle="Revenue in millions USD — all regions trending upward",
    source="Annual Report 2025",
)

Key Parameters

Parameter	Type	Description
x	str	Category column
y	str | list[str]	Value column(s)
group	str	Group column (long format)
orientation	str	"horizontal" (default) or "vertical"
mode	str	"grouped" or "stacked" for multi-series
sort	str | None	"descending" (default), "ascending", or None
limit	int	Show only top N categories
show_values	str	Value labels: "inside", "inside_end", or "outside"
reference_line	float	Draw a dashed reference line at this value

---

Scatter Plot

vizop.scatter(
    df, x="study_hours", y="exam_score",
    trend="linear",
    title="Study Time vs. Exam Performance",
    subtitle="Each dot represents one student",
    source="University registrar data",
)

Grouped scatter with highlight:

vizop.scatter(
    df, x="experience", y="salary", group="department",
    highlight="Engineering",
    title="Salary vs. Experience by Department",
    subtitle="Engineering salaries grow fastest with tenure",
    source="HR analytics, 2025",
)

Key Parameters

Parameter	Type	Description
x	str	Column for x-axis
y	str	Column for y-axis
group	str	Column for color-coded groups
size	str	Column for size encoding (bubble chart)
label	str	Column for point labels
trend	str	"linear", "lowess", or None
opacity	float	Point opacity (default 0.7)
jitter	bool	Add random noise to reduce overplotting
log_x / log_y	bool	Logarithmic axis scale

---

Slope Chart

Compare values between two time points. Supports both wide format (label/left/right) and long format (x/y/group):

df = pd.DataFrame({
    "country": ["Norway", "Germany", "USA", "China", "Brazil", "India"],
    "2015": [72, 65, 60, 38, 45, 28],
    "2025": [88, 78, 68, 62, 52, 48],
})

vizop.slope(
    df, label="country", left="2015", right="2025",
    highlight=["China", "India"],
    show_change=True,
    title="Renewable Energy Adoption",
    subtitle="Clean energy index score, 2015 vs. 2025",
    source="Global Energy Monitor",
)

Key Parameters

Parameter	Type	Description
label	str	Entity name column (wide format)
left / right	str	Value columns for each endpoint (wide format)
x / y / group	str	Long format alternative
highlight	str | list[str]	Entities to highlight
color_by_direction	bool | dict	Color by increase/decrease
show_change	bool	Show delta on right-side labels
sort	str	"ascending", "descending", or None

---

Waffle Chart

Proportional area chart using a grid of cells. Accepts a DataFrame or a simple dict:

vizop.waffle(
    values={"Employed": 62, "Unemployed": 5, "Not in labor force": 33},
    title="U.S. Labor Force Status",
    subtitle="Share of civilian population aged 16+",
    source="Bureau of Labor Statistics, Jan 2025",
)

Key Parameters

Parameter	Type	Description
values	dict[str, float]	Category-to-value mapping (dict mode)
category / value	str	Column names (DataFrame mode)
style	str	"square" (default), "circle", or "icon"
icon	str	Built-in icon name (required when style="icon")
grid_size	int	Cells per row/column (default 10 = 100 cells)

---

Raincloud Plot

Combines a half-violin density curve, box plot, and jittered strip plot for rich distribution visualization:

vizop.raincloud(
    df, value="score", group="treatment",
    title="Treatment Response Distribution",
    subtitle="Drug A shows higher median response with less variance",
    source="Clinical trial NCT-2025-0042",
)

Key Parameters

Parameter	Type	Description
value	str | list[str]	Value column (long format) or columns (wide format)
group	str	Group column (long format)
show_density	bool	Show half-violin (default True)
show_box	bool	Show box plot (default True)
show_points	bool	Show strip points (default True)
bandwidth	float	KDE bandwidth (default: Scott's rule)
jitter_width	float	Vertical jitter range (default 0.15)

---

Parliament Chart

Semicircular seat layout for legislative composition. Accepts a DataFrame or dict:

vizop.parliament(
    values={"Democrats": 213, "Republicans": 222},
    color_map={"Democrats": "#3571C1", "Republicans": "#D94444"},
    title="U.S. House of Representatives",
    subtitle="119th Congress — Republicans hold a slim majority",
    source="clerk.house.gov",
)

Key Parameters

Parameter	Type	Description
values	dict[str, int]	Party-to-seats mapping (dict mode)
party / seats	str	Column names (DataFrame mode)
rows	int	Number of arc rows (auto-calculated if None)
arc_degrees	float	Arc span in degrees (default 180)
majority_line	bool | int	Draw majority threshold (True = auto)
center_label	str | bool	Label at arc center (True = total seats)

---

Bump Chart

Rank chart showing how entities change position over time. Values are automatically ranked within each period:

vizop.bump(
    df, x="week", y="points", group="team",
    highlight=["Arsenal", "Man City"],
    title="Premier League Title Race",
    subtitle="Cumulative points through matchweek 8",
    source="premierleague.com",
)

Key Parameters

Parameter	Type	Description
x	str	Time period column (3+ unique values)
y	str	Value column (auto-ranked per period)
group	str	Entity column
highlight	str | list[str]	Entities to highlight
top_n	int	Show only top N by final rank
show_rank	bool	Append rank to endpoint labels (default True)
rank_order	str	"desc" (highest = rank 1) or "asc"

---

Configuration

Use vizop.configure() to set global defaults. Settings apply to all subsequent charts:

vizop.configure(
    accent_color="#E15759",
    background="light_gray",
    source_label="Finance Dept.",
)

Before (defaults):

After configure:

Setting	Type	Default	Description
accent_color	str	"#4E79A7"	Default color for single-series charts
font	str	"Inter"	Font family (bundled: Inter, Libre Franklin, Source Sans Pro, IBM Plex Sans)
background	str	"white"	"white" or "light_gray"
size	str	"standard"	"standard", "wide", or "tall"
source_label	str	None	Default source text for all charts
gridlines	bool	False	Show gridlines by default
dpi	int	300	Output resolution for .save()

---

The Chart Object

Every chart function returns a Chart object:

chart = vizop.line(df, x="date", y="value", title="My Chart")

chart.show()                    # Display interactively
chart.save("output.png")       # Save to file (default 300 DPI)
chart.save("output.png", dpi=150)  # Custom DPI
chart.to_base64()               # Base64-encoded PNG string
chart.close()                   # Close the matplotlib figure

In Jupyter notebooks, charts render inline automatically — no .show() needed.

---

Philosophy

vizop is opinionated by design. These defaults exist so you spend time on your data, not your chart settings:

• Titles are left-aligned to the plot edge, matching the natural reading direction of Western text. Centered titles are a vestige of Excel defaults.
• No legend boxes on line charts. Labels appear directly at line endpoints where your eye already lands. Legends force readers to bounce between the legend and the data.
• Bars sort descending by default. The most important value goes at the top. Alphabetical bar order almost never helps the reader.
• Horizontal gridlines only. Vertical gridlines add noise without aiding value estimation. The sole exception is scatter plots, where both axes carry continuous data.
• No top or right spines. They frame the chart like a box but add zero information. Removing them lets the data breathe.

---

License

MIT — see LICENSE.