PyPi package – https://pypi.org/project/tabpy/2.3.1/.
GitHub release – https://github.com/tableau/TabPy/releases/tag/2.3.1.
The most important improvement of this release is a bug fix for overriding deployed functions (models).
TabPy supports overriding for deployed functions with increasing endpoint version for it.
For example if you run TabPy locally and deploy function like in the example below you will have endpoint Fn version 1:
from tabpy.tabpy_tools.client import Client
client = Client('http://localhost:9004/')
def fn(x, y):
return x + y
client.deploy('Fn', fn, 'My awesome function')In case you want to make any changes in the function in the way users of it automatically pick up new version without any work on their side use override parameter when deploying again:
def fn(x, y):
return x * y + x * 2
client.deploy('Fn', fn, override=True)With overriding endpoint Fn its version will be increased and all user SCRIPT_... calculations from now on will call this new version.