Welcome to Datapane

High-level introduction to Datapane

What is Datapane?

Datapane is for people who analyse data in Python and want an easy way to share their results.

It provides a library which allows you to create reports programatically from components that wrap around the common objects in analyses: Pandas DataFrames, plots from Python visualisation libraries such as Bokeh and Altair, and Markdown text. Once created, reports can be published on the web, dynamically generated in the cloud, or embedded into your own application, where data can be explored, and visualisations can be used interactively.

Code
Report
Code
stocks.py
import altair as alt
import pandas as pd
import datapane as dp
df = pd.read_csv('https://query1.finance.yahoo.com/v7/finance/download/GOOG?period1=1553600505&period2=1585222905&interval=1d&events=history')
chart = alt.Chart(df).encode(x='Date', y='High', y2='Low').mark_area(opacity=0.5).interactive()
dp.Report(dp.Table(df['High']), dp.Plot(chart)).publish(name='stock_analysis')
Report

If you want your report to be generated dynamically by other people, for instance your team, you can deploy your Python script or notebook to a Datapane hosted server using Datapane's CLI. If you share it with others, they are able to provide parameters through a friendly web form, which are passed into your script.

Code
Web form
Code
stocks.ipynb
import altair as alt
import pandas as pd
import datapane as dp
ticker = dp.Params.get('ticker')
df = pd.read_csv(f"https://query1.finance.yahoo.com/v7/finance/download/{ticker}?period1=1553600505&period2=1585222905&interval=1d&events=history")
chart = alt.Chart(df).encode(x='Date', y='High', y2='Low').mark_area(opacity=0.5).interactive()
dp.Report(dp.Table(df['High']), dp.Plot(chart)).publish(name='stock_analysis')
CLI
$ datapane script deploy --script=script.ipynb --name=stock_analyser
name: stock_plot
# Script parameter definitions
parameters:
- name: ticker
type: string
default: GOOG
required: True
- name: start_date
type: date
- name: end_date
type: date
Web form

Our Mission

Although there are many enterprise BI and reporting tools with drag and drop interfaces, using SQL with Python is often the best combination for querying, analysing, and visualising data. Datapane's goal is to provide an API-first way to provide the last mile of sharing results, so you can analyse data in your existing environment, instead of using "Yet Another BI Platform".