Development Repository for Mach Perftest Notebook Tooling
Перейти к файлу
Gregory Mierzwinski 44b12c9949
Add archive note.
2024-03-15 08:29:47 -04:00
perftestnotebook Add Constant file. 2020-04-06 16:42:33 -04:00
testing Define a custom transformer with environment variable NOTEBOOK_PLUGIN. 2020-04-02 11:16:08 -04:00
.gitignore Update .gigignore 2020-03-19 17:46:49 -04:00
.travis.yml Define a custom transformer with environment variable NOTEBOOK_PLUGIN. 2020-04-02 11:16:08 -04:00
LICENSE Initial commit 2019-08-08 10:36:08 -04:00
README.md Add archive note. 2024-03-15 08:29:47 -04:00
__init__.py Add sample starter for mach-perftest-notebook. 2019-11-08 16:19:59 -05:00
requirements.txt Update requirements. 2020-03-19 23:34:39 -04:00
setup.py Update setpy.py 2020-03-26 10:28:22 -04:00
template_upload_file.html isort fixed imports. fixed minor formatting issues 2020-03-15 16:48:31 -04:00

README.md

This repository is not actively used so it's been archived.

mach-perftest-notebook-dev

Build Status

Development Repository for Mach Perftest Notebook Tooling. This tool will be used to standardize analysis/visualization techniques across Mozilla.

Running the tool

Install the tool with python3 setup.py install. Then call perftestnotebook --config=<CONFIG> with a config of your choice. See testing/configs for examples. This will standardize the data based on the Transformer you are using and output that. In the near future, it will also open an Iodide webpage to visualize and analyze the data with a set of pre-built analysis/visualization techniques.

Development Steps

These are the issues we will need to solve in this development repository (1 and 2 will likely need to be developed in parallel). The audience for this tool are software developers so we can take advantage of this for data transformation/standardization and other aspects of this tool.

  1. Develop data standardization techniques.
    1. Ideas:
      1. Build a decorator to decoarte custom transform/standardization functions in a provided script.
      2. Use a named function from a supplied script as the transform function.
      3. In addition to these two, build simple transformers for simple data formats so they don't need to be rewritten by developpers.
  2. Build some basic units of analysis/visualization in Python.
    1. Plotting data in a graph with custom x and y labels. Graph can be a line plot, scatter plot, or bar graph for now.
  3. Dynamically integrate basic units into an Iodide web page.
    1. Ensure that each basic unit can have it's standardized data modified easily.

Each of these steps should have proper testing for the functionality as well, pytest will work fine in this case.

Once something basic is built, we will look into builing the mach tool for this in mozilla-central.

Contributing

Feel free to ping us in #perftest on Riot for information on how to contribute.