Running process and method calling conventions
The Splunk platform applies logic when running process and method calling conventions. For MLTK and ML-SPL users, the following commands can be of particular interest:
- Running the fit command when partial_fit is False
- Running the fit command when partial_fit is True
- Running the apply command
- Running the summary commands
Running the fit command when partial_fit is False
When running the fit
command and the partial_fit
parameter is False, the fit
method of the chosen algorithm is called first. If that method returns a DataFrame the process returns to the search. If that method does not return a DataFrame, the apply
command is called to return a DataFrame.
The default for the partial_fit
parameter is False (partial_fit=f
).
Running the fit command when partial_fit is True
When running the fit
command and the partial_fit
parameter is True, the partial_fit
method of the chosen algorithm is called first on each chunk of 50,000 events. The apply
command is then called on those events. This process continues until you run out of events.
Running the apply command
When running the apply
command, the Python object is reconstructed using its codec. Then the apply
command is called on the events, chunk by chunk.
Running the summary command
Similar to running the apply
command, when running the summary
command, the Python object is reconstructed using its codec. Once done, the summary
command is called.
Custom algorithm template | Using codecs |
This documentation applies to the following versions of Splunk® Machine Learning Toolkit: 5.1.0, 5.2.0, 5.2.1, 5.2.2, 5.3.0, 5.3.1, 5.3.3, 5.4.0, 5.4.1, 5.4.2, 5.5.0
Feedback submitted, thanks!