Overview

If you've ever worked with DEX, you may come up with the perfect visualization or dashboard and want to carry it with you to other notebooks. This can be accomplished by using dx's built-in convenience functions.

Using pandas DataFrame .attrs

Set .attrs to customize your DEX display any time your pandas DataFrame is displayed.

df.attrs = {
    'noteable': {
        'decoration': {
            'title': "my super cool DEX dataframe"
        }
    }
}
df

Plotting

This is still very experimental. Any of the charts checked below indicate basic rendering should work (and allow manual user interactions in DEX), but not all combinations of plot arguments have been tested. Feedback welcome!

For official DEX chart documentation, visit https://docs.noteable.io/product-docs/visualize-data/dex-charts.

Supported Charts

• Basic Charts

  • Bar
  • Line
  • Pie
  • Scatterplot
  • Violin
  • Wordcloud
  • DataPrism

• Comparison Charts

  • Parallel Coordinates
  • Scatterplot
  • Connected Scatterplot
  • Scatterplot Matrix
  • Correlation Matrix
  • Bar
  • Dot Plot
  • Radar Plot
  • Diverging Bar

• Time Series Charts

  • Line
  • Cumulative
  • Stacked Area
  • Line Percent
  • Stacked Percent
  • Candlestick

• Relationship Charts

  • Force-directed Network
  • Sankey
  • Arc Diagram
  • Adjacency Matrix
  • Dendrogram

• Part-to-whole Charts

  • Pie
  • Donut
  • Sunburst
  • Treemap
  • Partition

• Funnel Charts

  • Funnel
  • Funnel Chart
  • Funnel Tree
  • Funnel Sunburst
  • Flow Diagram
  • Arc Flow

• Summary Charts

  • Big Number
  • Wordcloud
  • Dimension Matrix
  • Violin
  • Box Plot
  • Heat Map
  • Histogram
  • Ridgeline
  • Horizon
  • Hexbin

• Maps

  • Choropleth
  • Tilemap

Additional Functionality

• Dashboards
• Conditional Formatting
• Faceting

Enabling pandas plotting backend

To enable the dx plotting backend and use DataFrame.plot._____() or DataFrame.plot(kind=_____) syntax, you can run either of the following:

dx.enable_plotting_backend()

Or

pd.options.plotting.backend = "dx"

(They currently do the same thing, but dx.enable_plotting_backend() may handle more convenience functionality in the future.)

Warning

Not all chart options above can be called using the DataFrame.plot.______() pattern. Some examples:

• df.plot.violin(**kwargs) should instead be df.plot(kind='violin', **kwargs)
• df.plot.wordcloud(**kwargs) should instead be df.plot(kind='wordcloud', **kwargs)