Here's what's new in Release 1.0.9 (4 August 2017):
Bug fixes
Fixed an error triggered by running a display command on a DataFrame shared from a Scala notebook. PixieDust's scala magic feature allows users to share variables between Python and Scala Notebooks. Now, such shared Scala DataFrames are correctly displayed from within a Python context ( #425 ).
PixieApp chart options are now accepted as JSON. Instead of adapting parameters into HTML-style tag attributes, PixieApp developers can now simply paste valid JSON as the element content of the
<pd_options>
tag (
#423
).
Calls to
pixiedust.enableJobMonitor()
now work properly on Spark 2.1 (
#223
).
Fixed an issue where, within a PixieApp, the
showColumn
attribute would actually hide the column (
#383
).
PixieDust display calls to the map render for Google now work correctly when using Python 3.x ( #417 ).
Enhancements
The local installer experience has been improved. The
jupyter pixiedust install
command now allows users to more easily install and switch between different versions of the Spark kernel (
#348
).