Quickstart

CAVE is designed to adapt to any given workflow. While it’s possible to generate a HTML-report via the commandline on a given result-folder, CAVE may also run in interactive mode, running individual analysis-methods on demand. We provide a few examples to demonstrate this. Make sure you followed the installation details before starting.

Analyse existing results via the commandline

There are example toy results of all supported file formats in the folder examples on the github-repo. Run

cave examples/smac3/example_output/* --ta_exec_dir examples/smac3/ --output output/smac3_example

to run the example. By default, CAVE will execute all parts of the analysis. To disable certain (timeconsuming) parts of the analysis, please see the section commandline-options.

Most importantly though: CAVE will process one or several paths to directories with configurator output. glob -extension is supported. --ta_exec_dir defines a directory, from which the configurator was run - in case that there are relative paths while loading the data (e.g. instance-file-paths in SMAC’s scenario-file). Here also one or more values are valid, however either one path for all --folders-paths or exactly as many (one-to-one mapping). --output simply defines, where to save CAVE-output (report, plots, tables, etc.).

Interactive notebook mode

You can also run CAVE in an interactive notebook mode. Make sure you have jupyter installed, then just create a CAVE-object within a running notebook and run analysis-methods manually. See the jupyter-explanation for details. To run the smac3-example (within a notebook):

from cave.cavefacade import CAVE
cave = CAVE(folders=["examples/smac3/example_output/run_1"],
            output_dir="test_jupyter_smac",
            ta_exec_dir=["examples/smac3"],
           )
cave.parallel_coordinates()