arakawa 0.0.10

Build interactive reports in seconds using Python

Arakawa

Build interactive reports in seconds using Python.

Arakawa makes it simple to build interactive reports in seconds using Python.

Import Arakawa's Python library into your script or notebook and build reports programmatically by wrapping components such as:

• Pandas DataFrames
• Plots from Python visualization libraries such as Bokeh, Altair, Plotly, and Folium
• Markdown and text
• Files, such as images, PDFs, JSON data, etc.
• Interactive forms which run backend Python functions

Arakawa reports are interactive and can also contain pages, tabs, drop downs, and more. Once created, reports can be exported as HTML, shared as standalone files, or embedded into your own application, where your viewers can interact with your data and visualizations.

Getting Started

Check out our Quickstart to build a report in 3m.

Installing Arakawa

pip install arakawa
# or
conda install -c conda-forge

Examples

📊 Share plots, data, and more as reports

Create reports from pandas DataFrames, plots from your favorite libraries, and text.

import altair as alt
from vega_datasets import data
import arakawa as ar

df = data.iris()
fig = (
    alt.Chart(df)
    .mark_point()
    .encode(
        x="petalLength:Q",
        y="petalWidth:Q",
        color="species:N"
    )
)
view = ar.Blocks(
    ar.Plot(fig),
    ar.DataTable(df)
)
ar.save_report(view, path="simple_example.html")

🎛 Layout using interactive blocks

Add dropdowns, selects, grid, pages, and 10+ other interactive blocks.

...

view = ar.Blocks(
    ar.Formula("x^2 + y^2 = z^2"),
    ar.Group(
        ar.BigNumber(
            heading="Number of percentage points",
            value="84%",
            change="2%",
            is_upward_change=True
        ),
        ar.BigNumber(
            heading="Simple Statistic", value=100
        ), columns=2
    ),
    ar.Select(
        ar.Plot(fig, label="Chart"),
        ar.DataTable(df, label="Data")
    ),
)
ar.save_report(view, path="layout_example.html")

See the documentation for more details.

Acknowledgement

This project is fork of datapane/datapane and original codes are written by StackHut Limited (trading as Datapane).