Sciagraph is designed for profiling in production; as a result, it is only designed to run on Linux, on Python 3.7 or later.
As a result, to try out the profiling you will need to be running on Linux with Python 3.7 or later. If you’re using Windows, WSL might work. Or, on both macOS or Windows you can use Docker to start a Linux-based shell prompt:
docker run -it python:3.9 bash
In order to try Sciagraph out, you can install it in a virtualenv.
Make sure to upgrade
pip before installing Sciagraph, or you might have issues installing it.
Here are the commands to do so; substitute your version of Python if needed, e.g.
$ python3.9 -m venv ./test-venv $ . ./test-venv/bin/activate (test-venv) $ pip install --upgrade pip (test-venv) $ pip install sciagraph
See the documentation on setting up Sciagraph in production for detail on configuration for real-world usage.
You will need to set environment variables that enable usage of the Sciagraph profiler. To get an access key, sign up for a Sciagraph account.
$ export SCIAGRAPH_ACCESS_KEY=... $ export SCIAGRAPH_ACCESS_SECRET=...