Running the benchmarks

This document outlines how to setup and run the benchmarks using asv.

Running the benchmarks

To run the benchmarks with asv, simply enter:

asv run

Viewing the results

There are two steps to viewing the results locally. The results need to be published and the launched in a local web browser.

To publish the results use:

asv publish

And then to view the results:

asv preview

This will launch a local web browser from which you can view the results

Dry runs

One common use case is to use asv in development, there are several useful flags that should be used:

asv --python=same --quick --dry-run

We are adding three flags, these flags are:

1. --python=same uses the same environment as your development environment (saves time to avoid building environments)

2. --quick only runs the benchmarks once

3. --dry-run to not save the results of the benchmarks

These can be useful for getting quick feedback during development, but should not be used as anything other than a rough guide.