MaxCompute:Plotting

Prerequisites

Before you begin, ensure that you have:

• pandas and Matplotlib installed. Run pip install matplotlib to install Matplotlib and pandas.

• A Jupyter notebook connected to MaxCompute.

Quick start

The following example loads the pyodps_iris table and plots a single column as a line chart.

from odps.df import DataFrame

iris = DataFrame(o.get_table('pyodps_iris'))
%matplotlib inline

iris.sepalwidth.plot()

Plot examples

All examples use the iris DataFrame created in the quick start above.

Multi-line plot

Call plot() on the full DataFrame to plot all numeric columns as separate lines.

iris.plot()

Vertical bar plot

Use kind='bar' to produce a vertical bar plot. The example below groups by species, sums each group, and stacks the bars with a 30-degree label rotation.

iris.groupby('name').sum().plot(kind='bar', x='name', stacked=True, rot=30)

Histogram

Call hist() to generate a histogram grid with a shared x-axis.

iris.hist(sharex=True)

Supported chart types

Pass the kind parameter to plot() to select a chart type. For the full parameter reference, see pandas.DataFrame.plot.

plot() parameters

The plot() function supports the following parameters for axis and title customization.