We are happy to announce that we have now released Pycelonis 1.5.
Pycelonis is a python library built and managed by Celonis that allows easy interaction with Celonis’ pre-processed data as well as include functions to support building process-specific custom machine learning algorithms in python.
This release (Pycelonis 1.5) includes some major changes such as improving stability and robustness as well as bringing new functionalities. Following are the new changes and what it means for you:
New Push Data Functions
Pushing data to datapools is now split into 3 different functions covering 4 different behaviors:
Create new table in the datapool
data_pool.create_table(table_name="MY_PUSH_TABLE", df_or_path=df, if_exists="error")
Append data to an existing table in the datapool
Upsert data into an existing table in the datapool
data_pool.upsert_table(table_name="MY_PUSH_TABLE", df_or_path=upsert_df, primary_keys=['C'])
Replace data with the data in an existing table in the datapool
data_pool.create_table(table_name="Test_Data", df_or_path=replace_df, if_exists="drop")
You can then add the newly created table to the datamodel where needed using a different function which is:
data_model.add_table_from_pool(table_name="MY_NEW_TABLE", alias= "MY_NEW_TABLE")
What does this mean for you?
All the other functions in pycelonis that were previously available to push data into datapools or datamodel are now deprecated. This means that they will still work but we recommend switching to the new methods to have a better experience.
We have also added a new module in pycelonis that allows you to simulate process changes before implementing them.
If you want to learn more about it, check the following post on Process Simulation.
For more details and change log, please have a look at pycelonis documentation.
Celonis Product Team