Python setup

1. Ensure you have a proper Python environment. If you do not have the packages required in your default environment, consider creating a virtual one:

   python -m venv venv
   source venv/bin/activate
   pip install -r requirements.txt

2. Update your Python path

   export PYTHONPATH=`git rev-parse --show-toplevel`:$PYTHONPATH

3. (Optional) Set the Python log level:

   export PYTHONLOGLEVEL=info

   The default level is "warning", however most of the scripts produce useful information at "info". Valid values come from the Python logging module.

Assess response times

Parse raw data

From Google

Data can be parsed directly from a Google Sheet:

GOOGLE_API_KEY=... python src/parse-times.py \
	--sheet-id GSHEET_ID \
	--result-tab X > data.csv

where:

• GOOGLE_API_KEY is your Google Developer API key; this can also be specified as an option to parse-times.py (--google-api-key).
• GSHEET_ID is the Google Sheet ID
• X is the tab containing the data

Note that --result-tab can be specified more than once to combine data potentially spread across tabs.

From a local file

Data can also be parsed from a local CSV download of a Google Sheet:

python src/parse-times.py --data-file /path/to/data.csv > data.csv

In this case --result-tab is not required; however specifying which tab the CSV came from will make later plotting more readable.

Plot the results

python src/plot-times.py --output results.png < data.csv

where data.csv is the output of the data parsing script.

Some things to keep in mind:

• Call durations often have a long tail. Consider cutting off the graph at a reasonable duration to better view the distribution: using --cutoff 30 with src/plot-times.py, for example.
• The two steps -- parsing and then plotting -- have been presented separately for clarity. They can also be piped together to facilitate scripting:

  python src/parse-times.py ... | python src/plot-times.py ...