floris_metrics

Performance and test coverage metrics for FLORIS

Structure

This tool consists of a bash script and a Python script. The bash script lists all of the commits to include in the timing results and executes the Python script. The Python script defines a set of FLORIS cases, creates the input data, and prints the timing results.

sequenceDiagram
    get_quality.bash->quality_metrics.py: Run timing test
    loop Every commit hash
        get_quality.bash-->quality_metrics.py: Print timing results to terminal
    end

Loading

How to use

Software performance profiling can be tricky since modern operating systems typically have many processes running concurrently. This tool runs the timing tests a number of times and reports the average. The results can be used to understand the performance impacts of changes over commits. In FLORIS, this is used in a notebook that is included directly in the documentation.

To use with FLORIS, this directory must be parallel to the FLORIS directory:

.
├── floris
└── floris_metrics

Add more data points by listing new commits to the bottom of the list in the bash script. Then, run the bash script with this command from the parent directory of both FLORIS and FLORIS_Metrics:

>> $ ls
floris/           floris_metrics/
>> $ bash floris_metrics/get_quality.bash

The timing results are printed to the terminal window in a format similar to what is used in the FLORIS notebook linked above.