Vizflow

A data visualization pipeline tool for fast schema inspection, charting, dashboards, and export.

---

Vizflow turns raw CSV, JSON, Parquet, SQLite, SQL script, and piped stdin data into useful visual outputs without writing boilerplate notebooks. It profiles column types, recommends chart families, generates Plotly-powered interactive charts, serves local browser previews, combines charts into dashboards, converts datasets between common formats, and produces comparison reports for changing data files.

• Automatic schema inspection — detect semantic column types, null counts, unique values, samples, and numeric/date statistics.
• Smart chart suggestions — recommend line, bar, scatter, pie, heatmap, treemap, and histogram views from the data shape.
• Interactive Plotly charts — generate browser-ready HTML or static PNG/SVG/PDF exports.
• Dashboard mode — combine multiple chart specs into a responsive standalone HTML page.
• Local preview server — launch a Flask web server for chart, schema, and sample-data browsing on localhost.
• Data conversion and comparison — convert CSV/JSON/Parquet inputs and compare two datasets visually.
• Pipe support — stream CSV or JSON directly into commands with cat data.csv | vizflow plot.

Installation

pip install vizflow-cli

For local development:

git clone https://github.com/shazeus/vizflow-cli.git
cd vizflow-cli
pip install -e .

Usage

Inspect a dataset:

vizflow schema examples/sales.csv

Create an auto-selected chart:

vizflow plot examples/sales.csv --output sales-chart.html

Create a specific chart:

vizflow plot examples/sales.csv --type line --x date --y revenue --color region

Build a dashboard:

vizflow dashboard examples/sales.csv --charts "bar:region:revenue,line:date:revenue,pie:category:revenue"

Convert a file:

vizflow convert examples/sales.csv --to json --output sales.json

Preview in a browser:

vizflow serve examples/sales.csv --port 5050

Use stdin:

cat examples/sales.csv | vizflow plot --type bar --x region --y revenue --output piped.html

Commands

Command	Description	Example
vizflow plot <file>	Generate a single interactive chart. Omit <file> to read stdin.	vizflow plot data.csv --type scatter --x price --y volume
vizflow dashboard <file>	Combine multiple charts into one standalone HTML page.	vizflow dashboard data.csv --charts "bar:team,line:date:sales"
vizflow schema <file>	Inspect data types, nulls, unique values, stats, and chart suggestions.	vizflow schema data.json
vizflow convert <file>	Convert datasets between CSV, JSON, and Parquet.	vizflow convert data.csv --to parquet
vizflow serve <file>	Start a Flask preview server with chart and schema endpoints.	vizflow serve data.sqlite --table events
vizflow export <file>	Export a chart as PNG, SVG, HTML, or PDF.	vizflow export data.csv --format svg --type bar
vizflow compare <file1> <file2>	Compare two datasets and write an HTML report.	vizflow compare before.csv after.csv

Configuration

Vizflow is intentionally CLI-first and does not require a config file. Common options are available directly on commands:

Option	Purpose
--type	Choose auto, bar, line, scatter, pie, heatmap, treemap, or histogram.
--x, --y, --color	Override inferred chart columns.
--query	Run a SQL query against SQLite database or SQL script inputs.
--table	Select a table from SQLite database or SQL script inputs.
--output	Set the generated file path.
--format	Choose export format for chart output.

Static image export uses Plotly's Kaleido engine when available. If the local browser runtime is unavailable, Vizflow falls back to a Matplotlib renderer for PNG, SVG, and PDF outputs.

License

MIT License. See LICENSE.